EurWORK European Observatory of Working Life

Absence from work

Printer-friendly version
Paul Edwards and Kay Greasley

This study addresses patterns of absence from the 27 EU Member States and Norway, the costs involved, policies for dealing with absence and general developments in relation to promoting health and well-being. Average rates of absence across Europe are between 3% and 6% of working time. A reasonable estimate of the cost is about 2.5% of GDP. There is also evidence of presenteeism – the practice of attending work while ill – although there is a general lack of data on trends. While some countries are attempting to control costs, others put the emphasis on promoting well-being.

The study was compiled on the basis of individual national reports submitted by the EWCO correspondents. The text of each of these national reports is available below. The reports have not been edited or approved by the European Foundation for the Improvement of Living and Working Conditions. The national reports were drawn up in response to a questionnaire and should be read in conjunction with it.

Download the report (pdf, 203kb)

See also the executive summary


This study is based on a questionnaire distributed to the national correspondents of the European Working Conditions Observatory network. Its purpose is to show the overall extent of absence from work and outline the policies that have been developed in an attempt to deal with this issue, and to put this in the context of wider debates about the quality of work.

A study by the European Foundation for the Improvement of Living and Working Conditions (Eurofound) (Gründemann, 1997, p. 30) noted a lack of reliable data on the extent of absence. Recent work by the OECD (2009) estimates that the costs of disability and sickness benefits are 2.5 times those of unemployment, and that those costs are rising. An important question is whether modern data is more reliable and what is being done to manage attendance.

Promoting good health and attendance, instead of penalising absence, has become a growing policy issue. Presenteeism – the practice of going to work when ill or below par has also received attention. What is the balance between these two tendencies, and what do they tell us about the management of attendance?

Absence is defined as non-attendance at work when attendance was scheduled or clearly expected. An internationally-accepted definition of absenteeism is ‘the manifestation of a decision by an employee not to present themselves at their place of work at a time when it is planned by management that they should be in attendance’. The majority of absence is generally attributed to sickness or incapacity, but there may be other reasons. It is important to consider the pressures that lead to absence or attendance, to discover what level of illness justifies absence to an employee, if this view is shared by the employer; and what encouragement or pressure employers use to encourage attendance. In the context of recession, for example, does the fear of losing one’s job reduce absenteeism?

This definition can be only a starting point since, in practice, national data uses many different definitions. The extent of absence is hard to assess, and one objective of the study is to identify the basis of knowledge, and any key gaps. Similarly, the costs of absenteeism are inadequately quantified, and again the research aims to establish the basis and reliability of any estimates that exist.

This study draws on the contributions from the 27 EU Member States and Norway, comprising the network of the European Working Conditions Observatory (EWCO). Although these reports contain detailed information on sources, the study does not repeat their content; it attempts an overview and synthesis.

Definition of absence: sources and reliability of data

This study set out to analyse absences from work lasting three days or more. However, most national data sets make no distinction according to the length of an absence. Absences of any duration are thus the focus here. The generally accepted underlying concept is non-attendance when scheduled to work, so holidays and other planned leave are excluded (Huczynski and Fitzpatrick, 1989). National definitions vary, however, in terms of exactly what forms of absence are recorded. Table 1 summarises the bases of data and gives estimates of absence rates and trends. Precise definitions differ: for example, Norway excludes absence to care for family members, while Poland specifically includes it. In some countries, maternity leave is also included. Several countries, including France and Poland, have data that does not permit a meaningful statement of headline absence rates. In several other countries, estimates have been made based on stated numbers of sick days and estimates of the potential number of working days. In short, available statistics do not measure absence as defined above, and existing figures are calculated on many different bases.

Absenteeism statistics are generally derived from one of two sources: health insurance statistics, and surveys of employers or individuals. The former are comprehensive, but cover only the insured population and absences caused by sickness. In some countries, such as Greece, shorter absences are specifically excluded, while in many others it is likely that short voluntary absences are not reported because workers do not make insurance claims for them. Surveys include, in principle, all forms of absenteeism but depend on estimates that may be unreliable. Some national data excludes parts of the employed population, for example the public sector in Belgium. As there are sectoral differences in absence rates, such exclusions need to be taken carefully into account in making comparisons.

Stated rates of absence will vary between countries according to the type of information given. Thus the lowest rate of 0.8%, in Italy, comes from a household survey which asks for how many days household members took sick leave. Surveys of this kind are usually carried out over a particular period and only absences during this time are counted. Longer spells of absence may be missed or under-recorded. This is also true of, for example, the UK Labour Force Survey. A survey of self-reported absence in the Veneto region of Italy implies an absence rate of over 2%.

Clearly, making any direct comparisons between different national statistics would be unwise. As the French national report points out, comparisons within one country are difficult, primarily because of differences of definition and of the characteristics of the relevant populations. Making international comparisons is even more dangerous.

There have been some efforts to generate internationally comparable data – for example, a study of seven EU countries plus Canada and Switzerland, which drew on the Luxembourg Employment Study, itself based on labour force survey statistics (Barmby et al., 2002). Data was available in this study for only one year from each country. A few other studies exist, but the comparability of their data is unknown (e.g. Kaiser, 1996). More recent data covering more countries is not available.

Table 1. Absenteeism statistics: levels, coverage, trends, and sources
Country Headlineabsence rate Year ofdata Coverage Trends Data source Comments
AT 3.2% 2006 Absence of any duration Decline since 2001 Social security statistics Data include non-working days as well as working days. Some absences of up to three days may be omitted.
BE 5%–7% 2008 Absence of any duration Slight rise since 2002 Two surveys giving different estimates Excludes public sector.
BG 7.4% 2007 Absence of any duration Decline since 2002 National Statistics Institute; National Social Security Institute From 1 January 2007 onwards compensation payment of the sickness certificates changed.
CY Not stated   Claims for sickness benefit   Social Insurance Services Sick benefit paid to employed persons from fourth day of absence. Data relate only to sick pay claims.
CZ 5.2% 2008 Absence of any duration Decline since 2003 (but see comments) Business surveys Change in calculation methods in 2004 reduced reported levels.
DE 7.3 daysapprox = 3.2% 2008 Absence of any duration Slow decline since 2003 Medical certificates under health insurance Excludes privately insured and many part-time jobs. Published illness statistics give warning that they do not equate to the level of absence.
DK 4.6% 2007 Absence of any duration No clear trend Three surveys in different sectors (municipalities/regions, state and the private sector); Register of sickness benefits Data excludes enterprises with fewer than 10 employees. Several measurement and other changes prevent reliable estimates of trends. A comprehensive overview suggests constant absence levels over 30 years.
EE Not stated 2008 Absence of over one day No data Health Insurance Fund: benefits payments Reduced benefits paid from 1 July 2009: reported days absent likely to fall. Some under-reporting of diseases and accidents likely. Taking the stated number of days of sickness compensated, and dividing by the number of employed persons times an estimated number of days worked per annum gives a rate of 5.3%. Survey data indicates a rate of about 7%.
EL 4.7 days approx = 2% 2000 Absence longer than three days No data Social Insurance Institute Covers insured private sector workers only. Estimated absence rate based on a doctoral thesis; reliability of this estimate and comparability with other data not known.
ES 5.4% 2008 Absence of any duration Some increase 2004–2008 3 surveys of organizations Rate stated is from a survey of companies. Ministry of Work and Information survey data imply a much lower level. Comparability of these sources not stated.
FI 9.9 days approx = 4.3% 2008 Absence of any duration Some increase 2003-8 Labour force survey of individuals  
FR Not stated   Not stated Data on number of workers receiving sick pay show rise 1997–2004, followed by fall. Social security data; population surveys; organizational reports No standard measurement rules exist; data sources are not comparable.
HU 1.2%–1.3% 2007 Absence up to one year Decline 2003–2008 Health insurance fund Absence for first 15 days in any year paid for by employer, not the state.
IE 8 days approx = 3.5% 2007 Absence of any duration Lower than in 1970s Household survey; employer organisations surveys Household survey (the Central Statistics Office National Quarterly Household Survey) parallels the methodology of labour force surveys in other countries, such as Finland and the UK.
IT Approx 0.8% 2006 Absence of any duration Decline from a level of between 4% and 5% in 1970s Household and time use surveys Data sources very fragmented. Reported figure based on national centre calculations from Bank of Italy Survey of Household Income and Wealth. This may not be comparable with other data.
LT 7.35 daysapprox = 3.2% 2008 Absence of three days or more Increase since 2004 State Social Security Board  
LU 3.2% 2007 Absence of any duration No data Declarations by insured parties or employers  
LV Not available         Central Statistical Bureau reports that no data on patterns or distribution by sector, etc. is available.
MT 2.1 daysapprox = 0.9% 2008 Absence of 4 days or more Increase since 2004 Department of Social Security  
NL 4.1% 2008 Absence of any duration Decline from 2005 Netherlands Working Conditions Survey  
NO 7.7% 2009 Absence of any duration Decrease 2001–2008, followed by increase Survey of establishments; sick pay records Family illness and family and maternity leave are excluded.
PL Not stated 2007 Absence of any duration Stable 2003–20077, then increase Social Insurance Institute Includes absence to care for family members. Taking the stated number of days of sickness compensated and dividing by the number of employed persons times an estimate of number of days worked per annum gives a rate of 5.9%
PT 6.8% 2007 Absence of any duration Stable 2003–2007 Ministry of Labour and Social Solidarity, based on compulsory employer survey Definition includes all absence, whether justified or not. Maternity and paternity leave also included; excluding them would reduce the stated absence rate to about 6%.
RO 4.1 days; approx = 1.8% 2008 Not stated No data National Health Insurance Agency; labour force survey Labour force survey measures absence resulting from a disability caused or worsened by work; it does not measure sickness absence.
SE 3.1% 2008 Absence of any duration Steady decline from 2004 Labour force survey; social security data  
SI 3.7% 2008 Absence of any duration Decline since 2000 Institute of Public Health: insurance data  
SK 3.3% 2008 Absence of any duration Decline 2002–2004, then increase; the current level is below the peak in 2002 Social Insurance Agency Includes care for family members.
UK 3.3% 2009 Absence of any duration Slight decline from 2004 Employer surveys; labour force survey  

Note: Calculations in italics are by the authors.

Source: Contributions by national correspondents of the EWCO network; see detailed questionnaire returns for further comment and elaboration.

Absence levels and trends

Stated headline figures range from 0.8% in Italy to 7.7 % in Norway. Given the reporting differences noted above, this range may exaggerate differences. Eurofound’s earlier report (Gründemann, 1997, p. 18) gave a range from 3.5% to 8%. The range stated by Barmby et al. (2002) using a standard definition was from 1.8% to 6.3%. A range of 3% to 6% captures the central tendency in the present data. The mean absence rate for countries where a figure can be calculated is 3.8% with a median of 3.3%.

The research compared the mean absence rates in Table 1 with data from the Fourth European Working Conditions Survey conducted in 2005 (Parent-Thirion et al., 2007). This survey asked individual respondents how many ‘health-related leave days’ they had taken in the 12 months prior to the survey. Figure 1 shows the distribution of the 22 countries for which there was have reasonably comparable data on the two measures. There is a modest positive association (r = 0.485) between the two. Given the limitations of the statistical data on absence, this association is encouraging. It is also notable that in some countries, where data on absenteeism is not systematic, the association between the two measures is strong – for example, Romania. Countries where the absenteeism percentage gives a notably higher estimate than the number of leave days include Bulgaria, Norway and Spain. The reverse pattern is present in Finland, Malta, Slovenia, and Sweden. The countries do not cluster in any particular way: for example, Nordic countries appear in both groups. This may suggest that the two types of data are capturing a similar phenomenon, though more research would be needed to address this question.

Figure 1: Absence rate and self-reported leave days

Source: Parent-Thirion et al. (2007, p. 65). Headline rate of absence: Table 1 (above).

There are few clear patterns in the data. The Nordic countries used to be marked by high absence levels because of their generous welfare systems, with Sweden standing out in the Barmby et al. study. Norway’s high figure is consistent with this. Sweden’s low figure is thus notable and has possibly been affected by efforts to control sick pay costs since the early 1990s.

As noted above, national differences in absence rates may reflect differences in the structure of the working population. Thus, one of the lowest reported rates, in Malta, may reflect the low proportion of women in the work force. The Maltese national report notes that men here need to work as much as possible to attain the desired level of income for the family. A further possible factor is that women may be expected to have higher absence levels than men. As noted below, however, this is not the case in Malta so this factor is not operative.

There have, from time to time, been concerns about people going absent when they had no good reason to do so. Hard data on this question is always elusive, but the Dutch Working Conditions Survey – the main source of absence data in that country – finds that 94% of respondents denied taking time off work when they were not sick, and only 1% had done this for more than one day in a year.

Trends also vary between countries. Those reporting a decline slightly outnumber those reporting increases, but there is no common pattern. Such trends may result from two main factors: those focused on control, such as tightening sick pay regimes, and those focused on improving workers’ health. Hungary has also reported efforts to control sick pay costs, introducing rules in August 2009 to reduce sick pay and limit entitlements. However, there appear to be few studies that systematically test the nature of relationships in this area. Long-term patterns in Sweden may suggest an effect, but this may be distinctive given historically high levels of absence and generous welfare regimes; in other conditions similar effects may be lacking. Encouraging improvements in health and well-being may have a similar effect, for different reasons. This issue is discussed below, but such moves seem to be recent and to involve relatively few companies, so that overall effects on absence levels may be, as yet, hard to discern.

Factors that might increase absence levels include the age of the work force. France, for example, has an ageing work force, and here older workers generally report more absences than younger ones (see further below).

Patterns of absence

It is commonly argued that rates of absence for women are greater than those for men. The present data generally confirm this pattern, which was also found in the European Working Conditions Survey. There are marked differences in Belgium and Sweden (where stated rates for women are 60% more than those for men) and in Denmark, Slovenia and the UK (with a differential around 40%). In Estonia and Germany there are no marked differences, while Austria and Malta reported that rates for men are higher than they are for women. There is no clear explanation for this result, though it may reflect patterns of labour market participation. As for the latter, it has been already noted that Malta has a low percentage of female workers. It may be that women in the labour market are a distinct sub-group for example, those with a high commitment to paid work and thus low absence levels.

Two facts stand out. First, there is no uniform male–female differential. Second, the ordering of countries in terms of the differential does not seem to fall into any particular pattern. This suggests that it is the interaction of gender with other factors, such as the extent and nature of women’s labour force participation, the distribution of family obligations and the extent to which the social security system permits time off for family needs, and not gender itself, which is the important influence.

In relation to age, the general pattern is for older workers to be absent more than younger ones. However, the data in some countries, for example Estonia, shows that averages can mask important variations: older workers here are more likely to attend work, although if they do go absent it tends to be for a relatively long time. In Belgium, young workers have the greatest frequency of absence while older workers are, on average, absent for a longer duration. This is consistent with arguments going back to the 1950s, that young workers may use short absences as a form of escape from the demands of work, while older workers become accustomed to the demands of work, and go absent for health reasons (Hill and Trist, 1955). Such an interpretation might explain the UK data, where absence is reported to be higher for young workers. This reflects the fact that figures relate to reported absence during a given week: such figures capture any spell of absence and not the overall duration, and thus relatively understate the absences of older people. Portugal is the other country displaying this pattern; its data comes from employer surveys, which may also display a similar tendency.

The Hungarian data usefully consider gender and age together. Taking a reference category of men aged between 38 and 47 years, they show that younger men tend to have low levels of absence, while older men and women have higher levels. The highest rates are observed among women aged between 28 and 37, which is consistent with high pressures from childcare responsibilities in this group.

In several countries, long spells of absence account for a substantial proportion of the total days due to absence. Illustrative figures come from the following:

  • Austria: absences of over 14 days account for about 60% of the total;
  • Czech Republic: absences over 19 days account for 41% of the total;
  • Poland: absences over 11 days account for 44% of the total.

It is clear, however, that national conventions for classifying absence vary considerably, so that no directly comparable picture can be provided. The above estimates are also sometimes based on inexact information: for example, the distribution of days of absence according to the length of the spell of absence is not necessarily given. In the UK, for example, labour force survey reports cover absence during a particular week, and not the total length of absence. In several countries, there are no data on this issue.

A few countries relate absence to the size of the employing organisation. The general picture is one of low absence rates in small organisations. The Netherlands data is particularly clear on this point. In Finland, however, there seems to be little difference in this respect. The overall picture, also found in the European Working Conditions Survey, is consistent with evidence on small firms, which suggests that workers report a relatively high quality of work and also that sick pay provisions are less generous than they are in larger organisations (Edwards and Ram, 2009).

As for economic activity, results needed to be treated with caution: sector categorisation differs between countries, and in some cases only very broad categories are available. Seven countries have no data at all. There seems to be a pattern of absence being more prevalent in the public sector than it is in the private sector – for example, in Finland, the Netherlands and Sweden. Several factors have been suggested for these differences, including relatively generous sick pay arrangements in the public sector, and the fact that a good deal of public sector work is low-paid or stressful. This theme has been prominent in the UK, with several efforts being made to monitor and control public sector absence. The evidence here suggests that rates vary widely across the public sector with it not necessarily being a contributory factor to absenteeism. Indeed, absenteeism rates seem to be similar in the Czech Republic, Italy and Norway.

In relation to specific sectors, financial intermediation seems to have particularly low levels of absence, with eight countries (Austria, Bulgaria, the Czech Republic, Denmark, Finland, Norway, Slovakia, and the UK) reporting rates notably lower than the national average. The sector also stands out in this respect in the European Working Conditions Survey. In some countries, utility companies (electricity, gas and water) also report low levels of absence, though this is far from universal: the Netherlands, for example, reports a rate higher than average. There are no clear patterns for sectors with greater than average absences. In some countries, health and social work, and public administration, score relatively highly – for example, Finland and the UK. But this is not a uniform result.

Such differences are likely to reflect the composition of the work force and nationally specific variations. For example, high absence levels in public administration might be explained in terms of rising work intensity (Green, 2006), but in other countries there may be no such effect at work. It thus appears that absence rates are not strongly determined by the kind of employment involved.

Main causes of absence

The most common causes of absence are health problems. However, this study is mainly concerned with the management of attendance, rather than the specifics of health. Health conditions are, in fact, well covered in national surveys – for example, the detailed statistics on patterns of occupational diseases in the Czech Republic and Germany. Musculoskeletal and respiratory problems are very commonly identified as being among the top two causes. Back pain and syndromes such as repetitive strain injury also feature. A wide range of sources agree that musculoskeletal disorders ‘are the main occupational disease suffered by European workers’ (Eurofound, 2007, p. 2).

In some countries, broader issues such as monotony and work-related stress are mentioned in addition to physical conditions. Belgian surveys, unusually, quantify this: stress is found to be partly responsible for absence in about one third of cases.

Data in some countries, such as Denmark, reflect awareness that an absence can be sparked by many reasons. Thus, just how serious a back pain needs to be to lead to absence will depend on an employee’s commitment to the employer, the financial costs of going absent, and pressures to attend work. Models of attendance have long recognised such points (Steers and Rhodes, 1978), but they remain to be developed further. Simply knowing that respiratory problems are associated with absence says little about the social context that leads people to see such problems as a sufficient reason to take time off work. As discussed below, sick pay arrangements and health promotion schemes are important influences on such decisions.


‘Presenteeism’ has emerged as a distinct concept in the last 10 years. It is the practice of an employee’s attending work even when they feel too ill to be able to work effectively. It may be driven by a sense of loyalty to an employer or fellow workers, by compulsion, or both. How far is this practice recognised as a distinct phenomenon in Europe, and what is its incidence?

In 12 of the 28 countries, there are surveys or specific studies addressing the phenomenon. These surveys and studies differ in their depth and sophistication. Summary details can be found in the Annex

In addition to these 12 countries, in Slovakia the phenomenon is recognised although there are no data on it. In France, though no focused studies are reported, some companies have become aware of the issue and begun efforts to deal with it. Some countries, including Austria, the Czech Republic, Germany, Denmark and Finland, report substantial studies. The proportion of employees reporting that they had engaged in presenteeism is generally in the range of 50% to 70%, the highest being in the Netherlands, at about about 76%. Detailed academic studies are consistent with this. A Swedish study of the 1990s, for example, reports presenteeism among half its sample; it reports that the practice is most common among the caring professions and teaching (Aronsson et al., 2000). One Czech survey had a lower estimate, but added the interesting finding that employees may use their annual leavewhen they feel ill – arguably a variant on presenteeism.

Where reasons for presenteeism are stated, these tend to relate to a sense of duty to customers or colleagues. Fears or pressure to attend did not feature strongly, except in one Czech survey. Some studies cite negative effects on productivity: UK studies suggest that the costs may be greater than those of absence, though the evidence here is no more than suggestive.

The German data also suggests that presenteeism may be higher in small firms. This may be one explanation of the reported low levels of absence in such firms.

Studies of the effects of presenteeism seem to be rare. One study used the UK’s ‘Whitehall’ studies – a series of detailed analyses based on a large sample of civil servants (Kivimäki et al., 2005). It identified a group of male ‘sick presentees’: people who self-reported as unhealthy but who had no absence from work. This group had twice as many cases of coronary disease as people with similar health conditions who had been absent.

An academic overview adopts a different definition of presenteeism: broader, in that it covers any health problem at work, but narrower, in that it defines presenteeism as involving a reduction in performance (Schultz and Edington, 2007). It reports that research in the area is little developed and highlights a lack of validated methods to assess costs of the phenomenon.

Costs of absence

Gründemann (1997) emphasised the significance of absence costs in various countries, for example in the UK, Germany and Belgium. The report highlights the fact that the costs of absence could be reduced by tackling ill health. This, it is argued, will benefit employers, employees, government, insurance companies and society as a whole.

Table 2 gives details of how costs are calculated, together with estimates of the size of these costs. The table highlights how these costs are attributed to the employer, the nation and to specific social security budgets.

Considering the financial impact that absence from the workplace may have, it is notable that over half the national reports are unable to identify mechanisms for establishing the costs of absence. This is particularly common in some eastern European countries, such as Bulgaria, the Czech Republic, Estonia, Lithuania, Latvia, Romania and Slovakia. It is also evident in Denmark, Cyprus, Spain, Greece, Italy, Luxemburg, Malta, Netherlands, Portugal and Sweden. The lack of clarity on how cost statistics are compiled has certain implications. Firstly, it may mean that there is no available data on the cost to employers or governments. This is the case in Portugal, Slovakia, Luxemburg, Lithuania, Italy and Cyprus. Secondly, if the methods of computing costs are not given, it is extremely difficult to compare reported costs, as it is unclear what is or is not included in such estimates.

Where costs are recorded, the methods used are mixed. Costings are most commonly divided into direct and indirect. Direct costs may include the salary of the absent employee (or statutory sick pay), replacement costs and overtime costs. The direct costs can, in principle, be measured fairly clearly, though countries differ as to what is included. In the UK, for example, the social security system bears less of these costs than in other countries, and cost estimates from the UK should not be compared directly with those elsewhere. Belgium, however, classifies overtime as an indirect cost, again indicating differences in calculations of costs. The indirect costs may include the effects on productivity, administration, quality of service, social security contributions and the hiring of replacement workers; estimates here can also be highly variable. several case studies have been conducted of UK organisations, which find that it is rare for there to be any accurate means of estimating financial costs (e.g. Bevan et al., 2004).

The various methods used to calculate costs, plus the fact that several countries are unable to identify any methods at all, mean that any figures produced need to be considered cautiously. For example, in Malta the first three days of absence are not included in the cost calculation, and in Germany the figure does not consider all forms of employees. The differing methods in collecting data mean that it is difficult to compare the costs of absenteeism for each country. As noted in the introduction, the OECD has recently addressed this issue. It puts the cost of sickness benefits at 0.8% of GDP in 2005 across OECD countries (OECD, 2009, p. 36). Among the countries considered here, the range stated is from 5.4% in Estonia to zero in Portugal. As the national report on Portugal shows, while no data are reported, a substantial number of employees receive sickness benefit. Given estimates of absenteeism in Table 1, the costs may well be on a par with those of other countries. OECD data are also likely to omit many of the indirect costs. Where GDP figures are available, the average cost of absenteeism to a nation is 2.15% of GDP; there are no clear patterns or groupings of high or low GDP figures. Without clear definitions of what costs include and exclude, variances as a proportion of GDP levels are very difficult to compare. Absenteeism is clearly costly, but just how costly it is in any one country is hard to establish; with international comparisons particularly hazardous.

The approach to calculating the costs to social security systems is also variable, in terms of the the scope and detail of information recorded. Again, for a number of nations, no data are available. Social security payments are most commonly in the form of sickness benefit paid to those who are absent from the workplace. There are considerable national variations as to the number of eligible claimants for sick pay, and the amount of money they receive. In Finland, for example, the unemployed may receive payment, which is not the case in most other countries. In some countries the government plays a fundamental role in funding fundamental resourcing role – for example, Norway. However, in some countries, as in the UK, the employer takes primary responsibility for the costs of sickness absenteeism. Also notable is the role of insurance companies, which is evident in a number of EU countries, such as Austria, in the provision of insurance for those often on long-term sick leave.

The table also highlights the variance in costs to social security systems across the EU but, again, the methods used to calculate these costs make comparisons difficult. There does not appear to be a general trend as to the direction of these costs. Portugal reports a decline in the number of claimants whereas Cyprus reports a general upward trend. However, it is crucial to place these figures in context; for example, in Cyprus, an increase in employment figures has also occurred that would clearly affect the total costs. As the following sections will explore, there have been considerable developments in the area of absence management and health and wellbeing by both employers and governments. These changes may well affect the costs of absence not only for those directly bearing the costs but for employees too.

To sum up, the evidence suggests that there are considerable costs for several stakeholders when employees are absent and this is well recognised in some nations. The UK, for example, has a number of studies from employer organisations – the Chartered Institute of Personnel and Development (CIPD) and the Confederation of British Industry (CBI), along with AXA – government reports and media accounts of costs of absence to the public purse. However, for a significant proportion of countries, the interest in these figures is more muted, which begs certain questions. Firstly, is calculating the cost of absence too difficult? The difficulties of gaining reliable employer information and establishing the criteria for which costs are to be included can make this task an arduous one. The Swedish report notes that to calculate costs necessitates the inclusion of ‘innumerable variables’, many of which are inaccessible. Even when figures are provided, there is still a broad margin of error and many organisations are not included in the calculations (see the UK report for further details). Secondly, it is possible that absence is not seen as a primary concern and other areas are given higher priority. Ironically, it may be more likely that this area is a low priority because the full costs remain unclear. However, it is clear that, overall, the costs of absence are high, that measurement of these costs remains highly variable, and that direct comparison between countries is risky. That said, it may be that even the OECD underestimates the costs, partly because they assume zero costs in countries such as Portugal and partly because indirect costs are not calculated.

Table 2. Estimated costs of absence
Country Compilation of costs Estimated costs to employer, per annum National estimated costs, per annum Social security Comments
AT Direct costs; continued payment of wages. Indirect costs also include loss of production, overtime and administration costs Direct: €2.1bn Indirect: €3–4 billion Total: approximately €6 billion 2.2% of GDP Costs include: sickness benefits, treatment costs, accident insurance. Small companies also receive an allowance. Sickness benefits are paid by health insurance companies after a period of time - €375.91m(2003). There are also other large associated costs.  
BE Direct costs; average wage and employers labour costs Indirect; overtime, administration Direct; €3 billion Indirect; €3 billion x 2.5 Total; €10.5 billion (2008) 3.1% of GDP Three types of expenses: primary sick leave, maternity leave and disability benefits (when sick for more than a year). Costs for primary sick leave €998,675,055; 1“–2% of total social security budget (2007) Employer costs refer to private sector using simulated averages
BG No data No data 0.41% of GDP In 2009 the number of sickness absences increased by nearly 30% and the issue of work incapacity certificates increased by around 20%. GDP figure is total sum paid for temporary work incapacity. Some of the increase has been attributed to the recent use by certain companies of ‘fictitious incapacity certificates’.
CY No data No data No data The Ministry of Labour and Social Insurance annual report includes outlays by the Social Insurance fund: sickness benefits of €33.3 million, employment injury benefits of €2.1 million, which constitutes 3.7% of all expenditure by Social Insurance Services There has been a general upward trend but this is not considered of concern when employment level is taken into consideration
CZ No data available No data €1.7bn1.3% of GDP CZK 24.8bn was paid in sickness benefits (2002) (approximately €800 million) Sickness benefits excludes compensation related to accidents and occupational diseases as well as costs of medical care from accident and sickness insurance
DE Sick pay, maternity leave, social security contributions and health insurance contributions €34 billion €73bn (3% of GDP) Expenses of Statutory Health Insurance Scheme accounted for 4.11% ( €6.02 billion) of total costs of statutory health insurance. This cost has been in decline. National figure represents loss in gross value added. The figures do not represent all employees.
DK No data €25 billion No data Costs to sickness benefit were €1.61bn. No data available for Social Security System. The employers’ estimate does not include sickness absence insurance
EE No data No data 5.4% of GDP Incapacity for work benefits was EEK 2.4 billion (about €154 million), 62% were sick leave benefits. This accounts for 19% of the fund’s budget. Costs include benefits for sickness, nursing, occupational injury and maternity leave The GDP figure varies from 6-15% depending on methodology
EL No data €153 million – three fifths of the the annual spend of the Social Insurance Institute (IKA) and similar bodies No data Only figures provided by social insurance providers are available. Costs in 2001 were € 256 million, covering five days on average per insured person. The first three days’ wages are paid by employer Only a thesis from 2001 reporting data from 1998 is available. Analysed costs of IKA and other social insurance providers. Only examines those insured.
ES No data €2,500 per employee €12,700 million 1.21% of GDP No data Costs of absenteeism have risen 66% in previous five years.
FI Lost work input and premature retirement 5% of companies’ salary bill. One day of absence = €300 € 20 billion 16.7 million absence days covered by National Health Insurance. This is payable to unemployed also. However it does not include the first nine days of absence.  
FR Direct (regulation) costs: replacement, over time, excess costs. Indirect (disruption) costs: lower production, quality of service and loss of business No data to provide full account of costs to employers No data 5% of health spending Adjustments to absenteeism are impacted by the economic context and type of work organisation.
HU Data collected from enterprises on total costs of sick leave and costs of sick pay Sick leave: €21.5 per working day in 2008     Data from survey of labour related costs in enterprises. Published figures are hard to state as a proportion of GDP; no relative costs stated.
IE Indirect costs often excluded. Direct costs calculated using sick pay, overtime, replacement costs, etc. Several figures; €793 million – €1.1 billion, but estimates that cost could be as much as €2 billion Up to 1% of GDP 13,803 claims for occupational injury benefit were made. However this does not include absences of less than three days, the self-employed and certain public sector employments Non-inclusion of indirect costs could mean actual figure higher
IT No data No data No data No data No studies available
LT No data No data No data Sickness allowance accounts for 5.3% of State Social Insurance Fund Board (VSDFV) budget ( €193 million) Absence costs for employers are unavailable.
LU No data No data No data The relevant sickness fund makes payments for those insured from first day of illness. Estimated cost for blue collar employees - €196.6 million; Estimated cost for white collar workers - €25.0 million Blue-collar status no longer exists from 2009. Cash pay-outs have increased considerably, by 18.6% in 2006 and 13.6% in 2007. The deficit in 2008 was €18.5 million.
LV No data No data €57 million2.8% of GDP Costs include disability, maternity and sickness financing. Total costs were LVL 310.6 million (approximately €442 million); this figure is increasing.  
MT No data €35.97 million No data Figures from the Government Quarterly releases on Social Security Expenditure state that the cost of absence is €6.23 million; this figure has increased over the past five years. Figure may be underestimated as not all first three days of absence are recorded.
NL No data No clear data 2.96% of GNP (including prevention and work-related illness costs) Cost per employee as a result of work-related illness is €1,368. Cost of prevention per employee is €400. The social security costs listed also include costs to the employer. See national report for further details.
NO Indirect salaries such as holiday pay, social security contributions and other costs (estimated to be 50% of other costs) are included One day of absence costs around 3,000NOK (approximately €360) 80–100 billion NOK per year (approx. between €10.2 billion and €12.8 billion) No data Data is collected from the Sick Leave register and the Employee register, covering all doctor-certified sick leave. This does not include short-term absence. NOK 50-70bn (approx. between €6.4 billion and €8.9 billion) Costs of absenteeism are shared with employees covering 1/3 and social security covering 2/3.
PL Collected by social security administration. Remuneration paid to absent workers by employers PLN 3.42 billion (around €823 million) 0.29% of GDP Data collected by Social Insurance administration. Costs include sick benefits paid by the Social Insurance Fund.  
PT No data No data No data The number of claimants of sickness benefits is 549,891. This figure has declined. No estimates of studies available
RO No data €162.4 million is the employers’ contribution to the National Unique Fund for Health Insurance (FNUASS) No data Aggregate spending by FNUASS of €208 million. The first half of 2009 indicated a slight increase in this figure. The employers contribution has a shortfall of €55.2 million.
SE Calculation requires innumerable variables of which many are not accessible No data No data SEK 108bn (approximately €11 billion). The figure includes sick benefits and costs of rehabilitation. Although no data are available, costs assumed to be high for the social security system and higher still for employers.
SI Compensation of earnings paid by employers during sick leave of less than 30 days’ duration due to occupational disease or accident at work 1.4% of labour costs No data Expenditure for the compensation of earnings paid by compulsory insurance is closely linked to levels of absence. In 2006, €168,507,887 (9.1% share) which has declined since 2004. After 30 days, sickness benefit is paid by Health Insurance of Slovenia. Not all employees are included. Only absence due to occupational diseases or accidents at work is recorded.
SK No data No data No data The social insurance agency collects relevant data. Individual payers and the agency make contributions. Employers monitor number of days absent on a monthly basis but data is not published.
UK Direct: occupational and statutory sick pay; costs of replacement labour.Indirect: overtime, administration and reduced performance GBP 692 per employee (€841) GBP13 billion (€15.8 billion) 175 million days With indirect costs, estimated at GBP 27bn (€32.8 billion) Statutory sick pay is payable for absences lasting four days or more at a rate of GBP 79.15 per week (€96.2). This is payable by employers, except in special circumstances. The estimate given has a broad margin of error. Many organisations do not have detailed systems and only 41% of organisations calculate any costs.

Note: Figures in italic are estimates by the authors, taking stated total costs and dividing by Eurostat figures for GDP for 2007.

Source: Contributions by national correspondents of the EWCO network; see detailed questionnaire returns for further comment and elaboration.

Well-being at work

According to Eurofound, maintaining and promoting the health and well-being of workers comprises one of the four elements for improving the quality of work and employment (Eurofound, 2002, p. 6). Well-being has emerged in policy debates in recent years for several reasons. These include: the ageing of the workforce, together with concerns about the ability of people to work beyond the conventional retirement age; continuing concerns about productivity; and the European policy aim to improve quality of work.

This study has attempted to categorise initiatives in relation to well-being at national and social partner level into some broad categories. The categorisation is given in Table 3. Examples in relation to the most extensive developments are detailed in Table 4. Company-level examples are shown in Table 5.

As shown in Table 3, in nine countries clearly developed, formal, strategies exist for managing of employee health – strategies that include governments and social partners. One of the most extensive of these is in Norway, where a formal agreement has been in place since 2001. Austria has had developed policies since 2000. In Denmark, there has been an action plan to reduce absence since 2003. In 2008, a further tripartite agreement was reached, embracing means to allow the long-term sick to return to work gradually. The practice of employing people with reduced capacity to work on special terms (Skåne –og fleksjob) has developed. A 2007 survey (Virksomheders sociale engagement) reports that the majority of employers, and as many as 89% of those in the public sector, have policies in place to reduce absence and to keep the long-term sick in work. In other countries, such as Portugal, developments are more recent.

A further three countries – Bulgaria, Greece, and Slovakia – have strategies that are categorised as operating mainly at national level. They are.

In the UK, various approaches have been taken in the area of well-being policy but these have not been standardised into formal policies. There have been significant developments, featured – in particular – in the report ‘Working for a healthier tomorrow’. Many of its initiatives have been accepted by the government, and extensive well-being strategies are reported among employers. But, as in other policy areas, there is no formal set of agreements between government and social partners.

In four other countries, interest in well-being has emerged, but perhaps less clearly than in the UK. As an example of the recent emergence of well-being as an issue, in Slovenia the Health Insurance Institute of Slovenia (ZZZS) held a conference to promote good practice in absence management in April 2009. Among examples at company level, the pharmaceuticals firm Krka d.d. has a team project to improve working relations.

It is notable that the degree of intended effects on absenteeism varies. In Portugal, for example, the national strategy makes no direct reference to absence levels, and indeed absence from work is not central to policy debates. By contrast, reductions in sick leave are an objective in Austria, though this is seen as an ultimate result of a strategy whose focus is health and well-being, and not the sole or direct target.

Two countries are classed as reflecting some limited emergence. In Hungary, for example, well-being is not a feature of national debates, but there are examples of the implementation of programmes, mainly by large companies.

In nine countries, as listed in Table 3, well-being is not notably on the national agenda. In one of these countries, the Netherlands, a large survey of employers throws light on the balance of concerns: 17% reported ‘promoting a healthy lifestyle’, whereas 38% cited monetary or other incentives to reduce absence levels.

There appears, therefore, to be something of a bimodal pattern. Well-being features strongly in one group of nine countries, as listed in the first row of Table 3. It features very weakly in another nine – those in the bottom row of the table.

Table 3. State of development of well-being policies
Broad pattern National examples
Developed at national and social partner or company level Austria, Belgium, Czech Republic, Germany, Denmark, Finland, Ireland, Norway, Portugal
Mainly national strategy Bulgaria, Greece, Slovakia
Emergence, not institutionalised UK
Slow emergence Spain, France, Sweden, Slovenia
Some limited emergence Hungary, Lithuania
Not on agenda Cyprus, Estonia, Italy, Luxembourg, Latvia, Malta, Netherlands, Poland, Romania

Source: Contributions by national correspondents of the EWCO network

There is some evidence that approaches to absence in terms of well-being can have some effects. A Swedish study (Johansson, 2007; see EU0709029I) introduced the concept of ‘adjustment latitude’ to refer to the ability of people to adjust their work to their level of illness, so as to be able to retain sufficient capacity to work. It found that this sort of leeway tended to reduce the levels of absenteeism, and also to improve the likelihood of someone returning to work after being absent. This approach is consistent with developments in a number of countries. As the OECD (2009, p. 19) notes, in Denmark there was a ‘fundamental conceptual shift’ in reforms to the disability system in 2003; this involved a move towards a focus on a person’s capability to work, in particular the extent to which a person is able to carry out a subsidised job. The UK will move to a system of ‘wellness notes’, indicating ability to work in place of the traditional doctor’s sick note. Norway introduced in 2008 a system of avventende sykemeldin, which is best termed as partial sick leave, to cover cases where the person is able to keep working if some adjustments are made to the job (NO0809059I). Finland has had a scheme since 2007 that allows a worker absent for more than 60 days to return to work part-time and claim a partial sickness allowance.

The extent of social partner involvement in well-being naturally tends to follow the pattern outlined in Tables 3 and 4. Thus, clear engagement in national networks, together with a role for works councils, is reported in Austria and Belgium. At the other extreme, where well-being is not on the agenda, it follows that there is little direct social partner engagement. It is important however, to distinguish between well-being and the broader issue of health and safety at work. As Broughton (2008, pp. 17–19) notes, in Cyprus and Poland, for example, there is notable social partner engagement in wider debates, but these appear not to have developed in terms of well-being. One explanation in relation to Cyprus might be that the topic of absence is not seen as a central issue in national debates. As for Poland, employers place particular attention on the need to reduce the costs of absence, and this emphasis may prevent an approach in terms of well-being from developing strongly.

Within these broad patterns, the mechanisms of social partner engagement vary. Alongside countries where works councils appear to be the key mechanism, there are some where the law and collective bargaining play a combined role. This is the case in Slovakia, apparently with considerable effect. In other cases, collective bargaining is the main mechanism, with some impact, in terms of the inclusion of health and safety provisions in collective agreements, in the Czech Republic for example. In Ireland and the UK – as would be expected given the traditionally limited role of national agreements in governing the workplace in these countries – arrangements depend mainly on local practice.

Table 4. Examples of national and social partner initiatives on well-being
Country Initiatives
AT • Austrian Network on Workplace Health Promotion, begun 2000 • Safety and Health Management System • Draft bill to enshrine in law health promotion and well-being published 2009
BE • Law on the well-being of employees at work (1996) • Agreements between government and social partners
CZ • ‘Safe Enterprise’ programme, sponsored by state agencies and taken up by companies • ‘Company Supporting Health’ programme, supported by the Ministry of Health
DE • Common Occupational Safety Strategy, defined by the national conference of government bodies, accident insurers and social partners • 2009 Tax Act allows a tax rebate to employers for health improvement practices
DK • Modernised sick pay system and emphasis on prevention since 2003: see DK0401NU03 • Well-being a central aspect of debates: see DK0612039q • Set of 75 good practice cases covering public enterprises initiatives to improve the well-being of employees • Emphasis on ‘caring attendance’.
FI • Since 1998, four programmes involving government bodies and/or social partners have been launched
HU • Family Friendly Workplace Award of Ministry of Labour
IE • Health and well-being strategy by the Health and Safety Authority • Workplace Wellbeing Initiative, launched in 2007 by the largest employers’ body (IBEC)
NO • More inclusive work life (IA) agreement of 2001 between state, regional and local authorities and trade unions; renewed in 2005 • Several company level projects
PT • National Strategy for Health and Safety at Work (2008-12) • Action Plan for Hygiene and Health at Work 2008-2012, an initiative launched by the General Confederation of Portuguese Workers (CGTP-IN)

Source: Contributions by national correspondents of the EWCO network

Despite these patterns of engagement, there is little information about the concrete operation of the relevant mechanisms. The extent to which workplace representatives are actively engaged in well-being initiatives at workplace level has not been studied very much. This point is made explicitly in relation to the Danish situation, despite the fact that concepts of partnership are well-developed in this country.

Turning to company practice, a Belgian survey from 2008 indicates growing moves towards proactive management as well as cost control. In this survey, of 760 employers, 61% of respondents indicated that they had an attendance management policy going beyond cost control.

Examples listed in Table 5 come from the present national reports. Other studies provide further examples. In the UK, for example, two organisations (Stockton Borough Council and London Electricity) involve workers’ representatives in focusing on health and managing attendance (Broughton, 2008, p. 8).

Table 5. Company case studies illustrating well-being at work and attendance management
Country Company Details
Belgium Sappi Paper products manufacturing Devolution of responsibility to line management and health improvement campaigns Absence rate fell from 10% to 3%
Czech Republic Denso Mfg Auto air-conditioning manufacturing Company rehabilitation centre and physiotherapy services; healthy eating scheme; links to local health insurer
Finland Vaasan & Vaasan Bakery ‘Early support model’ for workers with health problems. Absence triggers lead to discussion with supervisor One percentage point reduction in absence reported
France Renault Auto manufacturing Measures include working time reductions, bans on after-hours working, support groups, and means to alert occupational health practitioners
Greece SC Johnson Hellas Work–life balance programme including flexible work schedules and health checks.
Hungary Unilever . MOL Group Food products manufacturing Multifaceted programme including, stress reduction, sports promotion and healthy eating Oil company Long-term health promotion plan to increase awareness of health and safety at the workplace including help in stopping smoking, stress management, healthy eating and sports
Ireland An Post GlaxoSmithKlein State post company Monitoring of long-term absence; use of occupational health advisers Pharmaceuticals manufacturing Measures include lifestyle benefits. Package is designed to meet needs of individual employees, e.g. employee assistance programme and monthly ‘well-being days’.
Italy Corneliani Clothing manufacturing company, characterised by repetitive work and short cycle times Project involving social partners and occupational health and research bodies launched in 1996–1997 to overcome the high level of absence. Greater involvement of workers in work design resulted in lower absence.
Norway Siemens Profitek Factory at Ålesund Agreement between company and elected employee representatives: attendance bonus scheme, with reported halving of absence rateElectrical and electronic manufacturer Programme to improve workplace environment and health, introduced via dialogue at every level of the company: 20% reduction in absence reported.
Portugal Supermarket chain TAP Portuguese Company of Hypermarkets Auchan: co-operation between company, representatives and health and safety service to promote health. Airline – 24-hour day care centre
Slovakia Three companies Examples of stress management and health promotion practices
Slovenia Two companies Krka Comprehensive risk assessment and health improvement programmes; absence rates reported to have fallen. Pharmaceuticals manufacturer – aim of better working environment and improved interpersonal relationships
UK Three companies Self-report results from companies – proactive health management reduces absence rates.

Source: Contributions by national correspondents of the EWCO network.

The examples given offer various illustrations of good practice, and in some cases effects in terms of absence rates are reported. It has long been argued, however, that there is little hard evidence showing a link between particular practices and outcomes (Nicholson, 1976; Wooden, 1988; Spurgeon, 2002). Controlled trials are lacking, and field studies find it difficult to show clear causal effects. It should also be noted that the cases that are reported tend to be those that appears to be successful or are unusual or distinctive in some way.

There are some examples of attempts to provide benchmarks for a ‘healthy workforce’. An increasing number of companies in Norway are reported to use, as an indicator, the proportion of workers who have not been absent for two years. The national report says that a figure of over 20% indicates a good working environment, with a figure falling to 15% indicating the need for improvement.

Overall, there is evidence of considerable (and growing) attention to the issue of well-being, together with targets and programmes promoting it. Although there is also some research on these, systematic evidence as to their effects reducing absence remains limited.

Practice in individual countries

A picture of overall policy on managing attendance can be gleaned from how individual social partners deal with the problem, and from the assessments by national centres.

Two groups of countries emerge in this analysis. In the first, the emphasis is on the promotion of workplace health. This group includes Austria, Belgium, Denmark, Finland, and Norway. Some of these countries have developed specific laws and policies to deal with the issue. In Finland, the Occupational Health Care Act is supported by several workplace health development programmes. In Norway, the Working Life Agreements of 2001 and 2005 focused on the quality of working life, with links to company practice and health improvement programmes. The effectiveness of well-being initiatives remains debatable and in Norway the effects of the agreements have yet to be evaluated. In Austria, addressing the personal and immediate causes of ill-health, such as smoking, has had some effect; but there has been relatively little attention paid to the wider issues of work organisation and working conditions. In Belgium, social partner agreements exist at regional level. Denmark has seen extensive debate on well-being (DK0612039Q). It is also notable that in some of these countries, such as Belgium, the trade unions have put particular emphasis on health promotion. For example, the managerial approach to reducing absenteeism at De Post/La Post was questioned by the unions who highlighted the role of working conditions in leading to absence.

Sweden is a variant on this pattern. Its approach is in terms of tightening the rules on claiming sick benefit, while also developing effective well-being programmes at company level. The issue of attendance management is hotly debated: employers welcome cost control while unions fear that reducing access to sick pay will damage workers’ health. Proposed changes to the Work Environment Act, involving reduced powers for safety representatives, exemplify these tensions (SE0907029I). One notable development here is the compulsory reporting, since 2005, by companies of sick leave figures in their annual reports.

Slovenia has also reduced sickness benefit levels, with some of the larger companies promoting well-being. In the second group of countries, generally from eastern Europe, the emphasis is on cost control, mainly through reductions in sick pay coverage and payment levels. This is specifically highlighted in Bulgaria, the Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland and Romania, but also in Luxembourg and Malta. In the Czech Republic, for example, there has been significant attention to cost reduction, notably through the ending of sickness benefit for the first three days of absence, a change formalised in an Act of 2009 (no. 261/2007 Coll.). The employer’s insurance premium for sickness has also been reduced. The point is made in the national report that such an approach increases pressures towards what might be termed forced, as opposed to voluntary, presenteeism.

In Estonia, too, the first three days of absence now (since July 2009) yield no sick pay, and concerns similar to those identified in the Czech report are reported. In Hungary, employees on sick leave now receive 70% of their average gross salary in sick pay, a reduction from the previous rate of 80%. In Lithuania, sickness benefit for the period from the third to the seventh day of absence has been cut from 80% to 40% of the wage. Extensive action against fraudulent claims for sickness benefit is also reported since the start of 2009. Slovakia has also reduced levels of sickness benefit since 2004, with some of the costs being shifted from the state to the employer. This has been associated with a marked fall in absence levels.

In many of these countries’ reports, employers’ concerns to control costs, and union worries about tightening control, are mentioned. In Poland, a campaign in 2008 led to inspectors contacting 250,000 people claiming sick pay, with about 10% of these claims found to be illegitimate. Similar action has been taken in Bulgaria. Here, in 2007, laws were introduced to tighten procedures over the issuing of sickness certificates. There are also penalties for doctors who issue false certificates, although it has so far proved impossible to demonstrate malpractice. The case of steel manufacturer Stomana Industry is revealing. The firm, which had a reported absenteeism rate of between 20% and 28%, appealed to the social security ministry for help because it believed some sickness claims were false. When this had no effect, the company introduced an attendance bonus, with reportedly dramatic effects in reducing absence levels.

Concerns regarding cost control are also apparent in Malta, despite the country’s low absence levels. According to the national centre, these concerns reflect general financial constraints. Two company-level examples are given: a firm that forbids overtime for those who have been absent, and a firm that randomly checks on absentees in their homes.

In Hungary, the decrease in the number of people taking sickness pay is partly seen as a consequence of the stricter regulation of sick pay entitlement, with a smaller allowance, introduced in August 2009.

National cost reduction policies do not necessarily determine employer policies. In the Czech Republic, for example, Denso Manufacturing has implemented well-being initiatives despite national efforts to cut the costs of sick pay.

The remaining countries fall between these two patterns or have mixed pictures. In Germany and the UK, for example, there is growing interest in well-being and prevention of absenteeism, as opposed to control, but systematic national-level moves seem less advanced than in some other countries. In Italy, well-being emerged as a recognised national issue in 2007, here absence not generally being seen as a large issue. There have, however, been efforts to control costs in the public sector by tightening up on benefit payments. In France, too, specific plans to address absenteeism are reported to be at an ‘embryonic’ stage. In Spain, the issue is recognised as an important cost to firms without its becoming central to the social partners’ policy agendas. A pointer to the mixed picture here is research by Esade Business School and Egarsat, a mutual insurance company, which – in 2008 – found that about half of all Spanish firms had some kind of mechanism or policy to control absence; of these, two-thirds stressed control and sanctions, and one-third more proactive attendance management.

Finally, Cyprus and Portugal appear to be distinctive in that there is not much attention at all to the issue. The reasons for this, particularly in light of the latter country’s relatively high levels of absence as shown in Table 1 and Figure 1, merit further attention.


The research has uncovered two fundamental points. The first is the limited extent of knowledge about the extent, causes and costs of absence. The second is a shift in policies of management and control.

On the first point, comparative studies note the extent of national variations in levels of absence, and problems with the comparability of available data: data scarcely exist in some countries and are patchy in others. On top of this, the different definitions and methods of measurement make international comparison hazardous. This contrasts with the phenomenon of unemployment, for example, for which statistical agencies have made major efforts to devise common definitions and measures,

Labour force surveys of individual employees represent one method of achieving comparable data, and their value might be addressed further. In analysing patterns, the research has followed the standard practice of considering mean (average) rates of absence. But averages hide a great deal of variation: within any one country, there is likely to be a wide distribution of patterns of absence. The distribution of absence, in particular the proportion of the total days of absence accounted for by spells of different lengths, merits further attention.

Turning to substantive developments, there appear to be considerable differences in levels of absence between countries, though these do not fall into neat patterns; former tendencies for Scandinavia to see high levels of absence, for example, seem to have disappeared. There has also been no overall trend in the rate of absence, though we might expect the recent recession to exert a downward effect.

In terms of the management of attendance, two broad trends are evident. The first relates to control. It is reflected in a growing concern with the costs of absence, together with policies of controlling these costs, notably those to health insurance systems. It may also be connected to presenteeism: if managements are too controlling, workers may feel forced to attend work when they are ill. It would, however, be wrong to exaggerate such a tendency. Coercive forms of attendance control seem to be rare. There are pressures towards costs control, but their effects are likely to be mediated by other factors.

‘Control’, moreover, is balanced by the second main development, towards an emphasis on health and well-being. As highlighted in Table 3, several countries have seen systematic efforts, at national and company level, to improve employee health.

The balance between these two patterns is hard to judge, though the depth with which well-being has become institutionalised in some countries suggests that the concept is effective. If this approach continues to gain ground, the nature of the debate about absenteeism may shift. Traditionally, the debate was about a phenomenon that was seen as costly and that could be managed through penalties or incentives to attend. Approaches using attendance bonuses and the like seem to have periods of popularity but to have little lasting impact.

An approach in terms of well-being is different. It focuses on the health of employees, rather than the particular effects of ill health in terms of absence from work. Several examples exist, from a wide range of countries. This is clearly consistent with interests in the quality of jobs, but it also implies a challenging agenda involving significant attention to work organisation nd the creation of a healthier work force. How far examples go beyond the few truly committed organisations remains in doubt, as does the balance of costs and benefits. The potential gains are considerable, though it is too early to see a major shift in organisations’ approaches to attendance management.

Paul Edwards and Kay Greasley, IRRU, University of Warwick/University of Lancaster


Aronsson, G., Gustafson K., and Dallner, M., ‘Sick but yet not at work: an empirical study of sickness presenteeism’, Journal of Epidemiology and Community Health, No. 54, 2000, pp. 502–09.

Barmby, T. A., Ercolani, M. G. and Treble, J. G., 2002, ‘Sickness absence: an international comparison’, Economic Journal, No. 112, 2002, pp. F315–31.

Bevan, S., Dench, S., Harper H. and Hayday, S., 2004, How employers manage absence, Department of Trade and Industry, Employment Relations Research Series, No. 25, 2004, available at

Broughton, A., Eurofound, 2008, Working conditions and social dialogue, 2008, available at /ef/search/node/comparative OR tn0710019s?oldIndex

Edwards, P. and Ram, M., ‘HRM in Small Firms’, in A. Wilkinson et al. (eds), The Sage handbook of human resource management, London, Sage, 2009.

Esade Business School and Egarsat, Egarsat barometer on attitudes and behaviours towards absence from work, 2008.

Eurofound, Quality of work and employment in Europe: Issues and challenges, Foundation paper no. 1, Luxembourg, Publications Office of the European Union, 2002.

Giaccone, M., Eurofound, Managing musculoskeletal disorders, Dublin, 2007, available at /ef/observatories/eurwork/erm/comparative-information/managing-musculoskeletal-disorders. At

Green, F., Demanding work, Princeton University Press, 2006.

Gründemann, R., Eurofound, Preventing absenteeism at the workplace (summary), Luxembourg, Publications Office of the European Union, 1997.

Hill, J. and Trist, E.L., ‘Changes in accidents and other absences with length of service’, Human Relations, No. 8, 1955, pp. 121–52.

Huczynski, A.A. and Fitzpatrick, M.J., 1989, Managing employee absence for a competitive edge, London, Pitman, 1989.

Johansson, G., 2007, The illness flexibility model and sickness absence, Karolinska Institute, Stockholm, 2007, available at

Kaiser, C. P., ‘Employee absence incentives in the welfare state’, Journal of Socio-economics, No. 25, 1996, pp. 89–103.

Kivimäki, M., Head, J., Ferrie, J.E, Hemingway, H., Shipley, M.J., Vahtera J. and Marmot, M.G., ‘Working while ill as a risk factor for serious coronary events: the Whitehall II Study’, American Journal of Public Health, No. 95, 2005, 98–102.

Nicholson, N., 1976, ‘Management sanctions and absence control’, Human Relations, No. 29, 1976, pp. 139–51.

OECD, 2009, Sickness, disability and work: Keeping on track in the economic downturn, Background paper, 2009, available at,3343,en_2649_34747_38887124_1_1_1_1,00.html

Parent-Thirion, A., Fernández Macías, E., Hurley, J. and Vermeylen, G., Eurofound, Fourth European Working Conditions Survey, Luxembourg, Publications Office of the European Union, 2007.

Schultz, A.B. and Edington, D.W., ‘Employee health and presenteeism: a systematic review’, Journal of Occupational Rehabilitation, No. 17, 2007, pp. 547–79.

Spurgeon, A., 2002, Managing attendance at work: an evidence-based review, British Occupational Health Research Foundation, 2002, available at

Steers, R. M. and Rhodes, S.R., ‘Major influences on employee attendance: a process model’, Journal of Applied Psychology, No. 63, 1978, pp. 391–407.

Wooden, M., 1988, The management of labour absence: an inventory of strategies and measures, Working paper no. 97, National Institute of Labour Studies, Flinders University.

Annex – Surveys and studies covering presenteeism
Country Study details
Austria Upper Austrian Survey covering 2010 employees, referred in Reif, Martin (no year given): Health care, health behaviour and health relevant areas of life from the perspective of Upper Austrians (in German, PDF)
Belgium BELSTRESS (1994–2003) and Research Group Work, Organisational and Personnel Psychology (WOPP) surveys; no data reported
Czech Republic Survey on ‘public opinion on medical care and absence from work’, 1997–2001
Germany Three detailed representative surveys of extent and effects of presenteeism: • telephone-based survey by Bertelsmann Foundation (Bertelsmann-Stiftung) Health Monitor (in German); • The research institute of health insurer AOK (WIdO) conducted telephone interviews of 2,000 employed statutory AOK members, aged between 16 and 65 years, in 2007; • The Initiative Health and Work (IGA) measured the productivity loss due to presenteeism through a survey of 2,000 employees (IGA-Barometer).
Denmark The ASUSI survey, 2004, looked into the prevalence of presenteeism and mapped the causes of sickness presence (for more details see ASUSI, Delprojekt: Sygenærvær, 2008, links in Danish). Apart from the ASUSI survey, only minor or specific surveys have been carried out – for example: • a survey undertaken on behalf of the Union of Commercial and Clerical Employees (HK), 2008; • Survey of Financial Services Union Denmark members (Finansforbundet).
Finland Several surveys, from between 2003 and 2008: • working life barometer survey conducted by Central Organisation of Finnish Trade Unions (SAK), 2008; • survey on the relationship between work and free time conducted by the Ministry of Labour – a postal inquiry covering 5,400 employees between autumn 2003 and spring 2004; • the Association of Finnish Local and Regional Authorities, the Commission for Local Authority Employers and the Local Government Pensions Institution implemented a survey, covering 5,000 municipal sector employees, on the topics of occupational well-being, changes in it and its promotion, 2006.
Italy Recent survey in bank industry, in the North-eastern regions of Italy, carried out by Fisac-Cgil (the union of the financial sector)
Netherlands TNO Cohort study among employees using the basic setup of the ‘regular’ cross-sectional NWCS/NEA, beginning 2008
Norway Level of Living Survey: Working Conditions 2006 ( Samordnet levekårsundersøkelse 2006: Arbeidsmiljø, 3.6Mb PDF, in Norwegian).
Poland Internet survey ‘Job satisfaction and pro-health attitudes’ (in Polish) carried out on a sample of managers and professionals (n=1,051) in 2007
Portugal Small, qualitative studies: • Martinez, Luís Fructuoso (2007) A esperança é a última a morrer? Capital psicológico positivo e presentismo (Hope is the last to die? Positive psychological capital and presenteeism) in Comportamento Organizacional e Gestão, Vol. 13, no. 1,37–54, available at • Laranjeira, Carlos A. (2009) O Contexto Organizacional e a Experiência de Estress: uma Perspectiva Integrada (An integrated perspective of the organisational context and experience of stress) in Ver. Salud pública, 11 (1): 123–133, 2009, available at (in Portuguese)
Sweden Two representative surveys: • Survey Arbetsmiljön from Arbetsmiljöverket, 2007, available at • Survey conducted by the Association for Company Medical Health Service, Föreningen Svensk Företagshälsovård (in Swedish), covering 5,000 individuals aged 20 to 67.

Source: Contributions by national correspondents of the EWCO network.

Useful? Interesting? Tell us what you think. Hide comments

Add new comment

Click to share this page to Facebook securely

Click to share this page to Twitter securely

Click to share this page to Google+ securely

Click to share this page to LinkedIn securely