european industrial relations observatory on-line Working time developments

Transcription

1 Seite 1 von 24 european industrial relations observatory on-line Search Site Map Contact Help..... Home > Country: EU.COUNTRIES Working time developments This review of developments in the length of working time in 2002 and 2003 finds that average collectively agreed weekly working time in the current European Union (plus Norway) remaine relatively stable at around 38 hours. Across the EU and Norway, agreed normal annual workin time averages around 1,700 hours. Of three sectors examined, agreed weekly hours are highest in metalworking, followed by local government and banking. Average collectively agreed paid annual leave entitlement rose slightly to 26.5 days in In the EU acceding an candidate countries, average collectively agreed working time is considerably higher, at nearly 40 hours a week and over 1,800 hours a year. Here we provide an overview of the duration of working time - as set by collective agreements and legislation - in the European Union, Norway and 10 acceding and candidate countries in 2003 (and 2002), based on contributions from the European Industrial Relations Observatory (EIRO) national centres. Setting the length of working time is still a key issue in industrial relations. While (as we will see below major general cuts in weekly working hours have been rare for some years (with the most notable exception being the introduction of a statutory 35-hour week in France), they remain on the agenda in many countries. Furthermore, less dramatic and smaller working time reductions continue to be negotiated, often as a quid pro quo for workers in exchange for forms of flexibility sought by employers. Achieving further reductions in working time is still a central demand for the trade union movement across Europe. The European Trade Union Confederation (ETUC) adopted a resolution at its 2003 congress (EU F) which commits it to 'continue to campaign for the 35-hour week and innovative working time reduction measures via collective agreements, combined, where necessary, with legislation, including a revision of the working time Directive to ensure that it covers all workers, order to achieve better reconciliation of working and family life, catering to the need for quality jobs a training while also taking account of employment and environmental aspects.' To take an example at European sectoral level, at its 2003 congress the European Metalworkers' Federation (EMF) adopted a work programme for which stresses that 'the organisation and determination of working time remains at the forefront of the collective bargaining agenda and is an important tool in the struggle against unemployment' EMF reaffirmed its commitment - set out in a 1998 working time charter (BE F) - to a working week of 35 hours, and to a common Europea minimum standard of 1,750 hours maximum normal contractual working time calculated on an annua basis (which translates as a 38-hour week) and an annual maximum of 100 hours' paid overtime. Furthermore, the action points adopted at its 2003 conference by UNI-Europa, the European regiona organisation of Union Network International (UNI), commit it to 'keeping the reduction of working time as one of its key demands'. As with other areas of labour statistics, there are many difficulties in making international comparison on the length of working time. Comparable data are not collected in all countries, while particular problems include: the existence of different ways of calculating working time, with annual, rather than weekly calculation increasingly common in some countries (TN S); the fact that working time reductions in some countries have been introduced through extra days off or cuts in annual hours, leaving the normal working week relatively unchanged; the increasing use of schemes whereby weekly hours may vary considerably around an average over a reference period; the treatment of part-time workers; and the differing roles of bargaining and legislation, with the latter having an impact on actual hou in some countries, but setting only a maximum 'safety net' in others.

2 Seite 2 von 24 Normal weekly working hours figures are also problematic in terms of comparing working time between countries as they do not take matters such as overtime (TN S) or the length of annual and other forms of leave into account. The aim here is to provide some broad general data on current developments, while pointing out the pitfalls involved in comparisons. For reasons of space, and because our objective is not to provide a statistical guide, we do not provide full definitions of how the figures are arrived at for each country, but merely call attention to the problems. The figures provided should be treated with extreme cautio and the various notes and explanations read with care. Average collectively agreed weekly hours Collective bargaining plays a key role in determining the duration of working time in most of the countries considered here (though to a lesser extent in some of the acceding and candidate countries). However, the nature of this role differs widely between the countries, with different bargaining levels (intersectoral, sectoral, company etc) playing different parts, and bargaining coverage varying considerably (though averaging around three-quarters of the workforce in the current EU, and about four out of 10 in the new Member States joining in 2004). Furthermore, the importance of bargaining differs considerably between sectors of the economy and groups of workers The relationship between bargaining and legislative provisions on working time also varies between countries. Figure 1 below sets out the average normal weekly working hours in 2003 for full-time workers as set by collective bargaining, across the whole economy, for the 26 countries examined. Figure 1. Average collectively agreed normal weekly hours, 2003

3 Seite 3 von 24 * 2000 figure, ** Average all countries, *** Average EU and Norway, **** Average all acceding and candidate countries, ***** Average of 8 new Member States joining EU in 2004, ****** Average expanded EU (23 Member States) Source: EIRO. Figure 1 should be read in conjunction with the following notes. Belgium: figure refers to the statutory working week set by the intersectoral collective agreement (this was cut to 38 from 39 in January 2003).

4 Seite 4 von 24 Bulgaria: figure refers to the normal week set in the great majority of agreements. Cyprus: figure from Pancyprian Federation of Labour (PEO). Estonia: as set by law. Finland: figure is an estimate, based on typical provisions of sectoral agreements. France: since 1 January 2002, normal weekly hours must, by law, be set at 35 hours in all companies (those with fewer than 20 employees have an exemption scheme relating to overtime). Germany: figure covers the whole of Germany; the figure for west Germany was 37.4 hours i both 2002 and 2003 and the figure for east Germany 39.1 hours in 2002 and 39.0 in 2003; data from the Institute for Economics and Social Science (Wirtschafts- und Sozialwissenschaftliches Institut, WSI) collective agreement archive. Greece: figure from National Statistical Service (ESYE). Hungary: collective agreements usually lay down the statutory weekly working hours of 40 hours (although the Labour Code allows collective agreements to stipulate a shorter period, in practice deviations from the statutory hours are relatively rare). Italy: figure represents the mid-range of agreements which provide for hours per week Latvia: as set by law. Luxembourg: figure is an estimate. Malta: while no statistical data exist, most collective agreements specify a normal weekly working time of 40 hours (according to Library and Information Unit, National Statistics Office) Netherlands: figure based on a sample of collective agreements. Norway: figure represents 'normal working hours' for employees covered by collective agreements; employees working shifts (eg in metalworking) or at nights (eg in local government healthcare) work fewer weekly hours. Poland: as set by law. Portugal: no data available since 2000; figure from Ministry of Labour and Solidarity's Department of Labour, Employment and Vocational Training Statistics (Departamento de Estatística do Trabalho, Emprego e Formação Profissional, DETEFP). Romania: as set by the tripartite 'single national collective agreement', which provides a minimum basic framework for employment conditions. Slovakia: figure refers to workers in single-shift systems - equivalent figure for workers in twoshift systems was 37.4 hours; data from Bratislava information system on working conditions. Slovenia: figure based on norm in sectoral collective agreements. Spain: figure calculated from Ministry of Labour and Social Affairs (MTAS) collective bargaining statistics on average agreed annual hours (1,762 in April 2003), on the assumptio of a six-day week and 274 working days a year. Sweden: figure from Mediation Authority (Medlingsinstitutet). UK: figures from Incomes Data Services (IDS) Study on 'Hours and holidays' 760 (October 2003). Figure 1 gives figures only for in most cases these are identical to those for The overall average weekly hours for the 15 current EU Member States plus Norway remained virtually unchanged at 38.0 (down from 38.1 in 2002). The only major change between 2002 and 2003 came Belgium, where the maximum normal week set by the national intersectoral collective agreement was cut from 39 hours to 38 from 1 January 2003 (many sectoral agreements already provided for a week of 38 hours or lower - see below). In 2003, the range of normal weekly agreed hours across the EU and Norway was five hours - ie between 35 hours (France) and 40 hours (Greece). However, 14 countries have a normal working week of between 37 and 39 hours inclusive. Looking at the five-year period , the average agreed normal weekly hours for the current EU and Norway has fallen from 38.6 to 38.0, and much of this change probably arises from the progressive, legislation-driven introduction of a 35-hour week in France, which was completed at the beginning of The overall picture is thus one of relatively static weekly hours, with major reductions very rare. As noted, the main exception is France, where the figure has fallen by four hour (or over 10%) over the five years. In Belgium, there has been a two-hour (or 5%) reduction during the period in the maximum normal week set by the national intersectoral agreement. The Luxembourg, Sweden and UK figures have fallen by around one hour (2.5%-3%). In Germany, the figure for the west of the country has remained static, while that for the east has fallen slightly (by 0.2 hours). Finland has probably seen some reduction, but the sources of data used have not been consistent over the period, and the same appears to be true in Portugal, though no new data are available since The figures have remained unchanged over the five-year period in Austria, Denmark, Greece, Ireland, Italy, the Netherlands, Norway and Spain (though here there has been a very slight downwar trend). Thus, with some exceptions, the early years of the 21st century have been one of relative (bu not total) standstill in average collectively agreed normal working time. In the majority of countries, most reductions have been slight overall (though there have been exceptions in specific sectors and companies), and often accompanied by new flexibility measures. However, it is still the case that the EU/Norway average has continued to creep downwards, albeit slowly, in recent years.

5 Seite 5 von 24 Looking in more detail at collective bargaining on working time in 2003, perhaps the most important event occurred in Germany. Here, the German Metalworkers' Union (Industriegewerkschaft Metall, IG Metall), which had pioneered the 35-hour week in western Germany, achieving it in metalworking in 1995, sought to extend it to eastern Germany. It succeeded in achieving an agreement to this effect i the eastern steel industry (DE N) but failed in eastern metalworking, despite four weeks of strike action (DE F), and the existing 38-hour week was maintained. This was IG Metall's firs defeat in a major dispute since the 1950s and arguably underlines the perception that the era of majo collectively agreed general reductions in working time is over (at least for now) in most EU countries. Flexibility, in various forms, now appears to be the main theme in bargaining on working time. For example, sectoral agreements introducing or expanding working time flexibility were signed in 2003 i countries such as France, Germany, Italy and Spain. An interesting example was found in Denmark, where new collective agreements in the finance and slaughterhouses/meat processing sectors introduced an innovative system of 'individual options', whereby employees can decide on the use of certain amount of the overall wage sum - choosing between higher pay, higher pension contributions or more time off - within the collective framework of the agreements (DK F). However, it should be noted that significant overall working time reductions continue to be achieved in some specific cases - for example, in April 2003, an agreement at Allied Irish Banks (AIB) introduced a 35- hour working week, which is relatively rare in Ireland (IE N). Working time reductions do remain on the agenda of many unions in the EU, especially in those countries with above-average agreed weekly hours. For example, in December 2003, the Greek General Confederation of Labour (GSEE) issued its demands for a new National General Collective Agreement to be signed in 2004, which included working time reductions in the direction of a 35-hour week (GR N). In the opening exchanges of the 2004 bargaining round in Sweden, industry trade unions have opposed employers' proposals to put an end to the progressive working time cuts, worth 0.5% of paybill a year, which have been continuing since 1998 (SE F). Turning to the acceding and candidate countries, it should first be noted that the active role of collective bargaining in setting normal weekly hours is relatively slight in many central and eastern European (CEE) countries. Thus, in countries such as Estonia, Latvia and Poland, collective agreements do not tend to deviate from the statutory normal hours (usually 40 hours). The average agreed weekly hours across the 10 countries stood at 39.7 in 2003, down from 39.8 in hours higher than in the current EU and Norway. The only changes from 2002 to 2003 were in Polan where the figure (based on the relevant legislation), fell from by one hour, and Slovakia, where there was a half-hour fall. In general, the 10 countries still have a 40-hour normal week, with the exception of Slovakia and Cyprus, where agreed hours are around the current EU average. Working time reductions do not appear to be very high on the agenda at present, though with some exceptions. In Bulgaria, for example, trade unions launched a demand in September 2003 for a gradual reduction o normal weekly working time from 40 to 35 hours (BG F). The initial response of all employers organisations was negative, while the Ministry of Labour and Social Policy did not want to intervene, preferring to leave debate on the issue to the social partners. In Hungary, negotiations on the reduction of working time were held in the tripartite National Interest Reconciliation Council (Országo Érdekegyezteto Tanács, OÉT) during 2003, but appeared to reach a deadlock (HU N) Adding the eight new countries joining the EU in 2004 considered here to the 15 present Member States produces an average agreed normal working week of 38.6 hours in 2003 (down from 38.7 in 2002). This is 0.6 hours higher than the current EU average. Statutory maximum working week and day Collective bargaining on the duration of working time takes place in all countries within the framework of statutory rules on maximum working times. In the EU and Norway, these should, at the least, respect the provisions of the 1993 Directive (93/104/EC) on certain aspects of the organisation of working time, which include a 48-hour maximum working week (on average over a reference period not exceeding four months), a minimum daily rest period of 11 hours and a daily hours limit of eight hours for night workers. The candidate countries should have implemented the terms of the Directive by the time that they join the EU. Table 1. Statutory maximum working week, 2003 Cyprus 48 Denmark 48 France 48

6 Seite 6 von 24 Germany* 48 Greece 48 Hungary 48 Ireland 48 Italy 48 Luxembourg 48 Malta 48 Netherlands 48 Portugal 48 Romania 48 UK 48 Austria 40 Estonia 40 Finland 40 Latvia 40 Norway 40 Poland 40 Slovakia 40 Slovenia 40 Spain 40 Sweden 40 Belgium 38 * No explicit weekly maximum in Germany, the 48-hour figure represents an average based on daily maximum rules. Source: EIRO. As table 1 above shows, the countries break down into two main groups - those which set their maximum weekly hours at the 48 hours specified in the EU working time Directive, and those which operate a rather lower limit of 40 hours (or 38 in Belgium). In the first group of 14 countries, the statutory maximum is in excess of average collectively agreed weekly working hours, and of actual o usual average weekly hours (see figure 5 and table 3 below) - it thus appears to operate essentially a a safety net (though the 48-hour figure often includes overtime - TN S). In the second group 12 countries, the statutory maximum is far closer to average agreed or actual/usual weekly hours (an identical to agreed hours in some cases), indicating a more active role for the law in governing working time (though again, overtime may not be included in this figure). The acceding countries experienced some change in this area in 2003, with a weekly maximum introduced for the first time in Malta, and the maximum reduced in Poland (by one hour) and Slovenia (by two hours). The maximum was also cut by one hour in Belgium. These statutory maximum figures may be exceeded in many countries, in the context of working time flexibility schemes allowing weekly hours to be varied around an average over a reference period (as permitted by the EU Directive) (TN S). To take a few examples: in Austria, weekly hours may be varied up to 50 over a reference period, by agreement, if an average 40-hour week is maintained; in Denmark, the 48-hour maximum must be observed on average within a period of four months; in Estonia and Slovakia, the average working week may be up to 48 hours over a four-month period, if overtime is included; in Finland, weekly hours may be varied over a 52-week reference period, if an average 40- hour week is maintained; in the Netherlands, the 48-hour maximum must be maintained over a 13-week reference period. If no agreement is reached between employer and trade union (or works council), statutory maximum hours are nine per day, but by agreement daily hours may be extended to 12, as long as average weekly hours do not exceed 60 over a four-week reference period (an do not exceed 48 over a 13-week period);

7 Seite 7 von 24 in Norway, average weekly hours may vary and be as high as 48, as long as the 40-hour maximum is maintained over over a reference period of up to one year. In some specific circumstances, the reference period may be extended; in Portugal, since 2003, weekly hours may be increased to 60 in some circumstances; in Spain, weekly hours may be higher if a 40-hour average is maintained over a reference period; and in the UK, weekly hours may exceed 48 as long as this average is maintained over a 17-week reference period All the countries considered here also have a form of statutory maximum working day, as set out in table 2 below - though this introduced as recently as 2003 in Malta and Slovenia. In Cyprus, Denmar Ireland, Italy and the UK there is no explicit maximum working day (except for night workers), but a 13-hour maximum (in most circumstances) can be derived from the application of the working time Directive's minimum 11-hour daily rest period. Once again, whether or not overtime is included differs between countries, and daily hours may often be higher in the context of working time flexibility schemes. For example: in Finland, the eight-hour limit does not include overtime; in Germany, daily hours may be extended up to 10, if an eight-hour average is maintained ov a 24-week reference period; in the Netherlands, as mentioned above, if no agreement is reached between employer and trade union (or works council) statutory maximum hours are nine per day, but by agreement daily hours may be extended to 12, as long as average weekly hours do not exceed 60 over a four-week reference period (and do not exceed 48 over a 13-week period); in Portugal, since 2003, daily hours may be increased to 12 in some circumstances; in Romania, daily hours may be higher if the eight-hour average is maintained; in Spain, daily hours may be higher if the nine-hour average is maintained over a reference period; and in Sweden, daily hours may be varied in certain circumstances. Table 2. Statutory maximum working day, 2003 Cyprus 13 Denmark 13 Ireland 13 Italy 13 Greece 13 UK 13 Malta 12.5 Hungary 12 Austria 10 France 10 Luxembourg 10 Portugal 10 Slovenia 10 Netherlands 9 Norway 9 Slovakia 9 Spain 9 Belgium 8 Bulgaria 8 Estonia 8 Finland 8 Germany 8 Latvia 8 Poland 8 Romania 8

8 Seite 8 von 24 Sweden 8 Source: EIRO. Collectively agreed weekly hours by sector Turning from the whole economy to individual sectors, we provide figures below for average normal weekly working hours for full-time workers as set by collective bargaining in sectors selected to represent manufacturing industry (metalworking), services (banking) and the public sector (local government). While these more specific figures are probably more accurate than the overall average figures given in the previous section, extreme caution is again advised in their use, and the notes under each figure should be read carefully - all the caveats which are noted above also apply to the sectoral statistics. Comparing the three sectors, in 2003 the highest average collectively agreed weekly hours in the EU and Norway were found in metalworking at 37.9 hours, followed by local government at 37.6 hours and the banking at 37.2 hours (the same ranking as in 2001 and 1999). In all three sectors, average hours were below the overall whole-economy average of 38.0 hours. In all three cases there were slight falls over the five-year period , due mainly to reductions in particular countries: France, Finland, the Netherlands in the case of metalworking; France, Belgium, Greece, Italy, Luxembourg and Spain in the case of banking; and France and Italy in the case of local government. This confirms that working time reductions (generally relatively small) persist in particular sectors and countries, resulting in a slight but steady fall in the overall average figures. With regard to the acceding and candidate countries, agreed weekly hours are above the EU/Norway average, though often no higher than those found in some of the current Member States with longer hours. The difference between the average figures for the EU/Norway and the acceding and candida countries varies between 2.1 hours in banking and 1.4 hours in metalworking. Metalworking Figure 2 below shows collectively agreed normal weekly hours in metalworking (based on the relevan sectoral collective agreements, except where specified otherwise in the notes under the figure). No data are available for Estonia, Latvia and Poland, while the metalworking industry as such does not exist in Malta. The figure provides data only for 2003, reflecting the fact that there was virtually no change from the 2002 figure in all countries, with the exception of Spain and the UK where there wer slight reductions of hours. In the EU and Norway, the average agreed weekly hours in metalworking stood at 37.9 hours in 2003 Since 1999, average hours have fallen from 38.4 hours, largely because of reductions in France, Finland, the Netherlands and, to a lesser extent, Spain and the UK (though the data sources used fo the EIRO figures have changed in some countries over the years). In 2003 in the EU and Norway, the longest weekly hours in metalworking (40) were found in Greece, Portugal and Sweden (though the Swedish figure is likely to be lower in practice, due to various working time reduction methods) and the shortest in France and western Germany (35). The range, a five hours, is identical to that found for overall average weekly hours across the whole economy. The overall ranking of the countries is similar to that for the whole-economy figures, though working hours in west German and Finnish metalworking are notably lower than the national average, while those in Italy, Portugal and Sweden are notably higher. Overall, the average agreed working week in metalworking, at 37.9 hours, is slightly below the overall average (38.0). Turning to the acceding and candidate countries, data are available for six countries, with average weekly hours standing at 39.3 hours in hours above the current EU/Norway average. Ther were no changes from Average weekly hours in Slovakia and Cyprus are around the average figure for the current EU, while they stand at 40 hours (ie over two hours higher than in the present EU) in the four other acceding and candidate countries. In Slovakia, weekly hours in metalworking ar somewhat below the whole-economy average, while in the other countries they are at or near this average. Considering all the countries examined which will be in the EU after May 2004, average weekly hours in metalworking stood at 38.1 in 2003, half an hour shorter than the whole-economy average, and 0.2 hours longer than in metalworking in the current EU.

9 Seite 9 von 24 EMF's proposed minimum standard of a 38-hour week maximum working week has now been achieved on average across the present EU Member States (plus Norway), though not in Greece, Portugal, Sweden, Ireland, Italy, Luxembourg, Austria and Spain (though the existence of various flexibility schemes and working time reductions implemented through time off rather than a cut in the basic working week means that no firm conclusions can be drawn on this point). It is near to being attained in the expanded EU, but some way off in most acceding and candidate countries. Figure 2. Collectively agreed normal weekly hours in metalworking, 2003 *2002 figure, ** Average all countries, *** Average EU and Norway, **** Average all acceding and

10 Seite 10 von 24 candidate countries, ***** Average expanded EU Source: EIRO. Figure 2 should be read in conjunction with the following notes. Belgium: figure is for industrial and artisanal metalworking - in subsector of bridge construction and structural metalworking, agreed working week was 37 hours. Bulgaria: 37 hours for employees with harmful working conditions. Germany: figure, from the WSI collective agreement archive, applies to west Germany - the figure for east Germany was 38 hours. Greece: 35 hours in shipbuilding. Italy: estimate based on annual figure from Bank of Italy (Banca D'Italia) of 1,730 hours (assuming 28 days annual leave and 12 public holidays). Netherlands: figure takes account of time off awarded for working time reduction - the basic figure is hours a week. Slovakia: 33.5 hours for those working in workplaces with a risk of exposure to carcinogens. Spain: figures refer to the whole industry sector and are calculated from Ministry of Labour and Social Affairs (MTAS) collective bargaining statistics on average agreed annual hours (1,758.1 in 2003), on the assumption of a six-day week and 274 working days a year. Sweden: collective agreements covering provided that a total of 1.5% of paybill could be taken as paid leave, contributions to a working time account or cash, but there are no data yet on how these 'working time deposits' have affected working time. UK: figure, from IDS (see note to figure 1 above), refers to 'engineering (including electronics Banking Figure 3 below shows collectively agreed normal weekly hours in banking (based on the relevant sectoral collective agreements, except where specified otherwise in the notes under the figure). No data are available for Estonia, Latvia and Poland. The figure again provides data only for 2003, as there was no change from the 2002 figure in all countries, with the exception of the UK, where there was a very slight rise (the figure used is an average of company settlements). Since 1999, average weekly hours in banking in the EU and Norway have fallen slightly from 37.6 to 37.2, largely because of reductions in France, Belgium, Greece (where there was a 1.3-hour cut in September 2002), Italy, Luxembourg and Spain, with hours largely static elsewhere. In 2003 in the EU and Norway, the longest weekly hours in banking (40) were found in Sweden and the shortest in Belgium, France and Portugal (35). The range, at five hours, is identical to that found for overall average weekly hours across the whole economy. Working hours in banking are often low than the national whole-economy average - very much so in Belgium, Greece and Portugal, and to a lesser extent in Finland, Italy, the Netherlands, Spain and the UK. However, hours in banking are notably higher than the national average in Germany and Sweden. Overall, the average agreed working week in banking, at 37.2 hours, is 0.8 hours below the overall average (38.0). With regard to the acceding and candidate countries, data are available for seven countries, with average weekly hours in banking standing at 39.3 hours in hours above the current EU/Norway average. There were no changes from Average weekly hours in Cyprus are at the average for the current EU, around 39 hours in Slovakia and Malta, and 40 hours (ie around three hours higher than in the present EU) in the four other acceding and candidate countries. In Cyprus and Malta, weekly hours in banking are around an hour below the whole-economy average, while in the other countries they are at or near this average. Including all the countries examined which will be EU Member States from May 2004, average weekl hours in banking stood at 37.6 in 2003, one hour shorter than the whole-economy average, and 0.4 hours longer than in banking in the current EU. Figure 3. Collectively agreed normal weekly hours in banking, 2003

11 Seite 11 von 24 * 2002 figure, ** Average all countries,*** Average EU and Norway, **** Average all acceding and candidate countries, ***** Average expanded EU Source: EIRO. Figure 3 should be read in conjunction with the following notes. Cyprus: normal weekly working time 35 hours during summer (May-August) and 37.5 hours during rest of year. Italy: estimate based on annual figure from Bank of Italy of 1,639 hours (assuming 28 days

12 Seite 12 von 24 annual leave and 12 public holidays); figure refers to credit and insurance sector, most recent specific figure for banking is for 2000 (estimated at 37.3 hours a week, based on annual figure of 1,649). Malta: estimate from Malta Union of Bank Employees. Romania: figure based on collective agreements of five of the most important banks. Spain: calculated from COMFIA trade union figure for agreed annual working time (1,700 hours), on the assumption of a six-day week and 274 working days a year. UK: figure, from IDS (see note to figure 1 above), refers to 'finance'. Local government Figure 4 below shows collectively agreed normal weekly hours in local government in 2003 (based o the relevant sectoral collective agreements, except where specified otherwise in the notes under the figure). No data are available for Estonia, Latvia and Poland. The only change from 2002 to 2003 wa in Slovakia, where working time in local government was regulated by law (setting a 40-hour week) until 2003, when the sector's first collective agreement introduced a 37.5-hour week. Legislation, rather than bargaining, still plays a part in setting weekly hours in local government in countries such as Austria, Portugal and Spain. Since 1999, average agreed weekly hours in local government in the EU and Norway have fallen slightly from 37.9 to 37.6, mainly due to reductions in France and Italy, with few changes elsewhere. In 2003 in the EU and Norway, the longest weekly hours in local government (40) were found in Austria, Greece, Luxembourg and Sweden and the shortest in Italy (33). The range, at seven hours, rather wider than found for overall average weekly hours across the whole economy. Working hours local government are significantly lower than the national whole-economy average in Italy and Portugal, and to a lesser extent in Belgium, the Netherlands, Spain and the UK. However, hours in local government are notably higher than the national average in Austria, Luxembourg and Sweden. Overall, the average agreed working week in local government, at 37.6 hours, is 0.4 hours below the overall average (38.0). With regard to the acceding and candidate countries, data are available for seven countries, with average weekly hours in local government standing at 39.3 hours in hours above the current EU/Norway average. The only change from 2002 to 2003 was in Slovakia (see above). Average weekly hours in Cyprus and Slovakia are just below the average for the current EU, while a 40-hour week is the norm elsewhere (ie 2.4 hours higher than in the present EU). In Cyprus and Slovakia, weekly hours in local government are below the whole-economy average, while in the othe countries they are at or near this average. Looking at the expanded post-may 2004 EU, average agreed weekly hours in local government stood at 38.0 in 2003, 0.6 hours shorter than the whole-economy average, and 0.4 hours longer than in loc government in the current EU. At its most recent general assembly in 2000, the European Federation of Public Service Unions (EPSU) set the 35-hour working week as a 'priority demand' for its affiliates. It is still some way off reaching this target on average across Europe, though the 35-hour week has been achieved in France, Portugal and Italy. Figure 4. Collectively agreed normal weekly hours in local government, 2003

13 Seite 13 von 24 * 2002 figure, ** Average all countries, *** Average EU and Norway, **** Average all acceding and candidate countries, ***** Average expanded EU Source: EIRO. Figure 4 should be read in conjunction with the following notes. Austria: usual figure for local government employees; working time fixed by law or decrees, a almost all public sector employees are excluded from right to conclude collective agreements Cyprus: normal weekly working time 36 hours during summer (June-August) and 38 hours

14 Seite 14 von 24 during rest of year. Finland: the figure given is for maximum rather than normal weekly hours. Germany: figure, from the WSI collective agreement archive, applies to west Germany - the figure for east Germany was 40 hours; local government covered by national agreement for public sector. Italy: estimate based on annual figure from Bank of Italy of 1,458 hours (assuming 28 days annual leave and 12 public holidays); figure refers to public administration generally. Malta: figure from Local Councils Association. Portugal: as set by law (Decree-Law no.187/88 of 27 of May 1988). Slovakia: hours for those working in two-shift systems and 35 for those working in three shift or continuous systems. Spain: as fixed by the state budget law. Sweden: blue-collar workers in the municipal healthcare sector have a 37-hour week. UK: figure from IDS (see note to figure 1 above). Actual/usual weekly working hours Some of the problems with data on collectively agreed normal weekly hours are avoided in statistics on actual weekly hours worked, typically measured in labour force surveys. These figures give a mor accurate impression of how many hours workers actually work in a given week, as they include facto such as overtime and absence. However, the national data on average actual weekly hours is often problematic for purposes of comparison, given differing definitions. Notably, some national surveys d not distinguish between full- and part-time workers. Therefore, in figure 5 below we give the Eurostat figures for usual hours worked per week by full-time employees, based on its 2002 labour force surveys, for the EU Member States, Norway and 12 acceding and candidate countries. Usual hours worked per week correspond to the number of hours the person normally works, including extra hour - paid or unpaid - normally worked (but excluding travel time to and from work and main meal breaks Figure 5. Usual hours worked per week, full-time employees, 2002

15 Seite 15 von 24 Source: Eurostat labour force survey. In the EU and Norway, the highest levels of usual hours worked by full-time employees are found in the UK (by some distance), Greece and Spain, and the lowest in Italy, Norway and France. Between 2001 and 2002, the average figure for EU 15 fell very slightly from 40.1 hours to 40. The average figure for the 10 acceding countries was 1.4 hours higher than that for the current EU in All acceding and candidate countries apart from Cyprus and Lithuania have usual hours above the current EU average. However, the UK has longer usual hours than any of the acceding and candidat countries, apart from Latvia.

16 Seite 16 von 24 In all countries, men's average usual hours are longer than women's. The average gender differentia is 2.2 hours a week (or around 6%) in both the current EU and the acceding and candidate countries and is three hours a week or more in the UK, Poland, Ireland and Italy, but less than one hour in Austria, Sweden, Bulgaria, Romania, Slovakia and Cyprus. Usual hours worked are in all countries higher than the average normal collectively agreed working week. Average usual weekly hours in the current EU were 40.0 in 2002, compared with average collectively agreed weekly working time of 38.0 hours (it is not possible to calculate the figure exactly for the acceding and candidate countries, but the gap seems to be similar to that in the present EU). The average usual working week is within one hour of the agreed normal week in Greece, Hungary, Malta, Ireland, Luxembourg, Italy and Norway. However, in seven countries usual hours exceed agreed hours by two hours or more - Cyprus, Denmark, Germany, France, Slovakia, Latvia and the UK. In the UK, there is a 6.1-hour gap between agreed and usual hours - arguably reflecting that country's long hours and overtime culture and the low coverage of collective bargaining. The 2002 Eurostat labour force surveys also give figures for the usual hours worked per week by par time employees, and these are provided in figure 6 below. Figure 6. Usual hours worked per week, part-time employees, 2002

17 Seite 17 von 24 Source: Eurostat labour force survey. The highest levels of usual hours worked by part-time employees in the current EU and Norway are found in Italy, France, Belgium and Sweden and the lowest in Ireland, Denmark, Spain and Germany Between 2001 and 2002, the average figure for the EU fell slightly from 19.8 hours to 19.7 hours. The average figure for the 10 acceding countries was 3.5 hours higher than that for the current EU in 200 All acceding and candidate countries apart from Slovenia have usual part-time hours above the current EU average, while usual part-time hours in Italy are above the average for acceding countries On average, female part-time workers' usual hours are longer than men's in the current EU (by 0.9

18 Seite 18 von 24 hours a week) but shorter than men's (by 1.1 hours) in the acceding countries. The countries are almost equally divided between countries where male part-timers work longer hours and those where female part-timers work longer hours. In the former countries, the greatest differentials (with male pa timers working on average over three hours more per week than female part-timers) are found in Italy Luxembourg and Romania. In the latter countries, the greatest differentials (with female part-timers working on average over four hours more per week than male part-timers) are found in Sweden and Denmark. There seems to be little relationship between the length of full-time and part-time hours, though they are both comparatively long in a number of acceding and candidate countries (especially Romania, Latvia, the Czech Republic, Slovakia and Poland). It is notable that some of the countries with the highest usual full-time hours are among those with the lowest part-time hours - notably the UK and Slovenia - while some of the countries with the lowest usual full-time hours are among those with the highest part-time hours - notably Italy, France and Belgium. As noted above, the national data on average actual weekly hours are problematic for purposes of comparison, given differing definitions. However, for the 16 countries where this is available, this information is provided for full-time workers in table 3 below (many of the 2003 figures are not yet available). Table 3. Average actual weekly working hours, full-time workers, 2002 and Country All Women Men All Women Men Austria* nd nd nd Belgium nd nd nd Bulgaria Cyprus nd nd nd Estonia nd nd nd Finland nd nd nd Ireland Latvia* nd nd nd Malta Netherlands* nd nd nd Norway Romania Slovakia Slovenia Spain nd nd nd nd UK Source: EIRO. * 2001 figures. The figures in table 3 should be read in conjunction with the following notes. Austria: figures from Chamber of Labour (Arbeitskammer); figure for men is an estimate. Belgium: figures from Statistics Belgium (Institut National de Statistique/Nationaal Instituut voor de Statistiek, INS/NIS) labour force survey. Bulgaria: figures from National Statistics Institute (NSI); 2003 figure for second quarter. Cyprus: figures from Statistical Service of Cyprus labour force survey. Estonia: figures from Estonian labour force survey. Finland: figures from the Statistics Finland labour force survey. Ireland: figures from the Central Statistics Office quarterly national household survey, for third quarter each year. Latvia: figures from Central Statistical Bureau (Centrala statistikas parvalde, CSP) labour forc survey for November. Malta: figures from National Statistics Office for June each year. Netherlands: figures, from Central Statistical Bureau (Centraal Bureau voor de Statistiek, CBS), include overtime. Norway: unofficial figures based on Statistics Norway labour force surveys (employees

19 Seite 19 von 24 temporarily absent from work not included). Romania: figures from Institute of National Statistics (Institutul National de Statistica, INS); 2003 figures for first half of year. Slovakia: figures, from Slovak Statistical Office (Štatistický úrad Slovenskej republiky, ŠÚ SR labour force survey, refer to fourth quarter of 2002 and second quarter of Slovenia: figures from Statistical Office of the Republic of Slovenia (Statisticni urad Republike Slovenije, SURS), labour force survey; 2003 figures for second quarter. Spain: figures calculated on basis of National Institute of Statistics (Instituto Nacional de Estadística, INE) labour force survey, second quarter. UK: figures from Office for National Statistics (ONS) labour force survey; 2003 figures are for June-August. For 15 countries, some figures are also available for the average actual weekly working hours of part time workers -see table 4 below (again, many of the 2003 figures are not yet available). It should be noted that the definition of part-time workers varies from country to country. Table 4. Average actual weekly working hours, part-time workers, 2002 and Country All Women Men All Women Men Austria* nd nd nd Belgium nd nd nd Bulgaria Cyprus nd nd nd Estonia nd nd nd Finland nd nd nd France nd nd Latvia* nd nd nd Malta Netherlands* nd nd nd Norway Romania Slovakia Spain nd nd nd nd UK Source: EIRO. * 2001 figures. The figures in table 4 should be read in conjunction with the following notes. Austria: figures from Chamber of Labour; figure for men is an estimate. Belgium: figures from INS/NIS labour force survey. Bulgaria: figures from NSI; 2003 figure for second quarter. Cyprus: figures from Statistical Service of Cyprus labour force survey. Estonia: figures from Estonian labour force survey. Finland: figure from Statistics Finland labour force survey. France: figures calculated on basis of National Institute of Statistics and Economic Studies (Institut national de la statistique et des études économiques, INSEE) labour force survey; 2003 figure for first quarter. Latvia: figures from CSP labour force survey, for November. Malta: figures from National Statistics Office for June each year. Netherlands: figures, from CBS, include overtime. Norway: unofficial figures on basis of Statistics Norway labour force surveys (employees temporarily absent from work not included). Romania: figures from INS; 2003 figures for first half of year. Slovakia: figures, from ŠÚ SR labour force survey, refer to fourth quarter of 2002 and second quarter of Spain: figures calculated on basis of INE labour force survey, second quarter. UK: figures from ONS labour force survey; 2003 figures are for June-August.

20 Seite 20 von 24 National labour force statistics in Hungary, Ireland, Italy, Poland and Sweden do not distinguish between the average actual weekly hours of part-time and full-time workers, while French statistics d so only partially. Table 5 below gives the average actual weekly hours of all workers in these countries. Table 5. Average actual weekly working hours, all workers, 2002 and Country All Women Men All Women Men France nd nd Hungary* nd nd nd Ireland Italy nd nd nd Poland Sweden nd nd nd Source: EIRO. *2000 figures. The figures in table 5 should be read in conjunction with the following notes. France: figures refer to usual average weekly working time declared by employees in INSEE labour force survey; 2003 figure for first quarter. Hungary: figures from ILO; share of part-time employees is relatively low, so figures approximately represent hours of full-time employees Ireland: figures from Central Statistical Office (CSO) household survey for third quarter each year. Italy: figures from National Institute of Statistics (Istituto Nazionale di Statistica, Istat). Poland: figures, 'for average number of hours spent in regular and additional job', from Centra Statistical Office (Glówny Urzad Statystyczny, GUS); figures refer to third quarter of each yea Sweden: figures from Statistics Sweden labour market survey.. Annual leave The annual duration of working time is strongly influenced by the amount of paid annual leave to which workers are entitled. Figure 7 gives the average number of days of collective agreed annual leave for 16 countries where data are available (harmonised on the basis of a five-day working week The figures generally apply to 2003, with the only significant changes from 2002 being a one-day increase in Denmark. The average entitlement across the EU and Norway stands at 26.5 days, and has increased slightly over the past few years from 25.6 days in 2000 (though the number of countrie included and the calculation methods have not been uniform over this period). Agreed annual leave entitlement varies considerably, from 33 days in Sweden to 20 days in Ireland. Little information is available on this point from the acceding and candidate countries, with figures only for Cyprus and Romania - both of which are below the EU/Norway average. In most acceding countries, annual leav is probably at the statutory level (see below), though in Slovakia in 2003, in about 25% of organisations the collective agreement provided for additional leave over the statutory minimum, usually of one week. Figure 7. Average collectively agreed annual paid leave (in days), 2003

21 Seite 21 von 24 * 2001 figure; ** 2002 figure, *** Average EU and Norway, **** Average all countries Source: EIRO. Figure 7 should be read in conjunction with the following notes. Austria: expressed as 30 working days, including Saturdays; figure from Chamber of Labour Denmark: made up of five weeks annual leave, plus four days to be taken at employees' own choice or paid instead in 2002, rising to five from Finland: 25 days (or 30 days including Saturdays) applies after one year's service, 20 days (o

22 Seite 22 von days including Saturdays) for employees with less service. Germany: figure covers whole of Germany; the figure for west Germany was 29.2 days in bo 2002 and 2003; the figures for east Germany were 28.4 days in 2002 and 28.3 days in 2003; data from the WSI collective agreement archive. Greece: mid-range of entitlements which vary from 20 to 25 days, depending on service. Italy: figure calculated as four weeks' leave, plus the mid-range of between five-10 days awarded as a form of working time reduction. Luxembourg: figure is an estimate. Netherlands: figure represents 25.3 days of holiday, plus six days awarded in context of reduction of working time. Portugal: 24 or 25 days is the agreed norm, with the 25th day sometimes dependent on factors such as age and attendance. Romania: calculated on basis of tripartite single national collective agreement. Sweden: figure calculated as statutory 25 days, plus the mid-range of between five and 10 days additional leave awarded in most collective agreements. UK: figure from IDS (see note to figure 1 above). All countries examined here have a statutory minimum period of paid annual leave, as set out in figur 8 below. In the figure, the leave is expressed in days and harmonised on the basis of a five-day working week, and the statistics given are the basic entitlement (entitlement increases with length of service in some countries). The EU countries and Norway fall into two groups - those which have a 20-day minimum entitlement (21 in the case of Norway and 22 in the cases of Portugal and Spain), a laid down in the EU working time Directive, and a smaller group with a 25-day minimum entitlement. All acceding and candidate countries except Malta have a 20-day minimum entitlement. This 20-day figure was introduced in 2003 in Poland, Romania and Slovenia, all of which formerly had an 18-day minimum entitlement Average collectively agreed annual leave exceeds the statutory minimum by four or more days in Denmark, Finland, Germany, Italy, the Netherlands, Norway, Romania, Sweden and the UK - indicating that the law acts essentially as a safety net in these countries. Average collectively agreed annual leave and the statutory minimum are close or the same in Austria, France and Ireland - and also probably in most acceding and candidate countries - indicating a more active role for the law. Figure 8. Statutory minimum annual paid leave (in days), 2003

23 Seite 23 von 24 * Average acceding and candidate countries, ** Average all countries Source: EIRO. Figure 8 should be read in conjunction with the following notes. Austria: entitlement expressed as 30 working days, including Saturdays, increases by five days after 25 years' service. Cyprus: 25 days for those working a six-day week. Estonia: expressed as 28 calendar days - certain groups (such as people with disabilities)