Low-wage workers and the 'working poor'
Low pay continues to be a feature of employment throughout Europe, with around one in seven employees in the EU estimated to have a low wage - though definitions vary considerably. Furthermore, there has been increasing attention in recent years to the phenomenon of the 'working poor'. This EIRO comparative study looks at the extent of low-paid employment and working poverty across the EU and Norway, highlighting the different definitions used, and examines the factors that lead to low pay. It then sets the issue in its industrial relations context, examining the role of minimum pay systems, set by law or collective agreement, and of trade union bargaining strength, in influencing the extent of low pay.
In recent years, a debate has intensified across Europe over the issues of 'low-wage workers' and the 'working poor'. There has been considerable change and reform in European labour markets since the beginning of the 1980s, during which period the distribution of wages is widely regarded as having become more unequal. This is seen by some commentators as a result of: a weakening of workers' bargaining power under the pressure of unemployment; institutional change; and changes in the composition of the labour force. This increase in earnings inequality raises concerns about the consequences for the lives of workers located at the lower end of the wage distribution, with a growing proportion of workers receiving a wage so low as to harm their ability to maintain decent living standards. The response from some quarters is to call for the improvement or introduction of mechanisms to prevent excessively low pay, such as 'wage floors' of some kind.
From another perspective, other researchers (mostly mainstream economists) often argue that low wages reflect differences in the productivity of workers and that 'wage floors' obstruct the fundamental role of the market - ie the efficient allocation of resources. Low levels of productivity in many activities or occupations, usually in the service sector, require a low level of real wages in order to generate a rate of profitability comparable to that in high-productivity activities or occupations. Therefore, statutory minimum wages are considered as impediments to the entry of less qualified workers into employment.
The economic situation over the past 20 years in Europe has also generated a relatively new social category known as the 'working poor'- ie those who are in employment but remain below a defined poverty threshold. A common view is that the two situations - ie being a low-wage worker and a members of the working poor - are closely related, and that reducing the incidence of low pay would lead to less poverty.
The EU context
The issue of low pay has received a degree of attention from the European Union over the years. The 1989 Community Charter of the Fundamental Social Rights of Workers (the 'Social Charter') stated that 'all employment shall be fairly remunerated. To this end, in accordance with arrangements applying in each country, workers shall be assured of an equitable wage ie a wage sufficient to enable them to have a decent standard of living.' In 1993, as part of the action programme implementing the Social Charter, the European Commission adopted an Opinion on an equitable wage, in which it stated that the pursuit of an equitable wage must be seen as part of the general drive to achieve higher productivity and employment creation, and to foster good relations between the two sides of industry. The Member States were encouraged to work towards the establishment of an equitable wages policy, through greater labour market transparency with regard to wages. The social partners were also called upon to contribute to the achievement of this aim.
In 1997, the Commission published a progress report on equitable wages since its 1993 Opinion (EU9702103N). It concluded that most Member States have the basic legislative framework in place aimed at achieving an equitable wage. However: few Member States had viewed the Commission Opinion as a catalyst for action; the majority of Member States saw direct intervention in wage policy as undesirable; labour market developments, such as the growth of temporary and part-time employment, had restricted governments' capacity to influence wages policy; the data available indicated that the distribution of wages had generally become wider over time; wage inequalities had widened in a number of Member States; and transparency of wage information remained a problem.
The specific issue of low or inequitable pay has arguably not since featured prominently in EU policy. Instead, the focus in this area has shifted more towards tackling poverty and social exclusion and promoting the 'quality of work'. In 2000, the European Council made the fight against poverty and exclusion a key part of its new 'Lisbon' modernisation approach (EU0004241F). A social exclusion strategy was launched, with the Member States drawing up National Action Plans - submitted in 2001 - in response to agreed EU objectives (EU0111101N). The strategy stresses that the best safeguard against social exclusion is a job, and its objectives include a 'guarantee that everyone has the resources necessary to live in accordance with human dignity' and 'ensuring that the take-up of employment results in increased income'. However, there is little of direct relevance to low pay levels.
In terms of quality of work, in June 2001, the Commission issued a Communication on Employment and social policies: a framework for investing in quality, which takes forward the commitment in its current five-year social policy agenda (EU0007266F) to promote quality in employment and social policy. It proposed a set of possible indicators for quality of work, which include indicators related to 'intrinsic job quality'. The indicators suggested by the Commission to assess the achievement of the aim that 'jobs are intrinsically satisfying, compatible with persons' skills and abilities, and provide appropriate levels of income', include the number of low-wage earners (employees earning less than 60% of median income) and 'working poor' and the distribution of incomes, as well as the proportion of workers advancing to higher-paid employment over time. The Commission observes that: 'The relation between low or no skills and job prospects, low pay and poverty is important. A shortage of highly skilled people tends to push up wages at the top end of the income scale, while a surplus of people with few or no skills tends to push down wages at the bottom. The result is to increase the scale of social transfers needed in order to bring low-wage household incomes up to minimum, socially accepted, levels.' The Communication's approach was endorsed at the European Council summit meeting in Laeken in December 2001 (EU0201231N).
Finally, outside the EU framework, the Council of Europe's European Social Charter provides for the right to 'a fair remuneration', referring to 'a remuneration such as will give them and their families a decent standard of living'. On this basis, the committee of independent experts which advises on the implementation of the Charter recommended in 1977 a minimum 'decency threshold' of 68% of the average national wage. More recently, the committee has adjusted this threshold in the light of changing circumstances (eg the development of earnings patterns within families), stating that the lowest net wage should not fall below 60% of the net average wage.
The issues of low-wage workers and the working poor are topical at present, while - as mentioned above - there is considerable disagreement among researchers and experts in a range of disciplines (including economics and industrial relations) as to current developments in this area and their meaning. In this context, this comparative study - based on the contributions of the European Industrial Relations Observatory (EIRO) national centres in the EU Member States and Norway - examines:
- the proportion of total employment made up by low-wage workers - ie the groups in the labour force who are more affected by wage inequality - and the characteristics of these workers and their employment;
- the incidence of the phenomenon of the 'working poor' and its relation to low pay;
- the relationship between low wages and low productivity and the importance of 'wage floors'– especially minimum wage systems – in reducing the number of low-wage workers; and
- earnings mobility - ie the possibility for low-wage workers to move from the lower end of the wage distribution to upper levels.
The study thus:
- looks at the concept and definitions of low-wage workers and working poor in the countries examined;
- considers the level and general development of low wages across Europe, and the level of low pay;
- provides an overview of factors affecting the incidence of low pay;
- outlines the evidence on the progress of low-paid workers into higher-paid positions;
- examines the relationship between low-paid workers and the working poor;
- analyses the influence that institutional arrangements have on low wages - focusing on minimum wage arrangements, collective bargaining, trade unions, wage floors and labour market policies; and
- reviews the views and policies of the social partners in this area;
An annex to this study examines the coverage and quality of the relevant wage statistics.
As we will see below, the definitions, data sources and statistical methodologies used in different countries are not always identical and comparisons between countries cannot always be rigorous. Nevertheless, a number of overall trends can be identified.
Concept and definitions
Although there are common elements in the different definitions of low-wage workers used, mainly for policy, research and statistical purposes, across the EU and Norway, a strict definition of the concept accepted by all parties does not exist. Even if it is generally considered that low wages should be defined as wages below a threshold designating what is socially acceptable remuneration, the difficulty of determining what is socially acceptable in an objective way leads researchers and policy-makers to adopt different thresholds expressed as a proportion of the median wage - or sometimes average wage - of all workers.
The Organisation for Economic Cooperation and Development (OECD) defines low pay as less than two-thirds of median earnings for all full-time workers, and this threshold seems to have gained acceptance in research and statistics in countries such as Austria, France, Greece, Ireland, Italy and Portugal. Sometimes, 50% of median earnings, rather than two-thirds, is considered as the threshold for defining very low wages. The Eurostat definition of low pay as a monthly wage less than 60% of the median salary seems to be prevalent in Denmark. Researchers and interest groups in other countries use their own thresholds. For example, in Germany, a widespread definition of low pay is wages below 75% of the national average, while in Norway the definition most often used by trade unions and Statistics Norway is hourly pay below 85% of average hourly pay in manufacturing industry. The UK Office for National Statistics (ONS) has recently begun to use 'low paid' to mean those paid below the statutory national minimum wage. It seems that in some countries, collective agreements, institutions and tradition lead to an implicit definition of low wages as being related to some kind of minimum wage or minimum income.
Another divergence in definitions relates to the wage to be considered in the definition of low pay. It is argued in some quarters that the definition must apply to annual pay considered as an income, not as a cost. With the development of part-time work, the resulting growing number of low-paid workers (in overall income terms) cannot be captured by calculations based on hourly gross wage rates. Nevertheless, in some cases (such as Norway and some Irish research) the definition used is hourly pay; this means that low-wage workers are defined as those with hourly pay below a certain percentage of the average or the median hourly pay in the country considered. It seems that this definition, which does not fit well to the new realities of labour markets, persists in some countries because it was established through collective agreements and is embedded in institutions and tradition. A monthly reference point is used by Eurostat and in countries such as Austria, Denmark, France (where researchers also use annual figures) and Greece.
A further point of divergence between definitions relates to working time. In some cases - such as the widely accepted OECD definition - the definition applies to full-time workers (for reasons related to the uncertainties surrounding working time). Nevertheless, excluding part-time workers from the calculation of the number of low-wage workers underestimates the incidence of low pay. There is widespread agreement that the definition applies to all employees regardless of sector or gender. It is not always clear, however, if the definition of low-wage workers applies to employees who are apprentices or young workers in training (such workers are excluded in France, for example).
Official statistics provide no information on low pay relating to work in the 'underground' or 'clandestine' economy, usually performed by immigrants, young people and women. It is thus clear that low-paid work is severely underestimated, since it is likely to be more frequent in the underground economy. The extent of the under-reporting is proportional to the size of each country's 'non-official' economic activity.
The progress of research on low wages, and the number of studies and publications in this area, differs dramatically between countries. In some countries (such as Greece, Luxembourg, Portugal and Spain) there is almost no research, or it is only beginning, whether in other countries (such as France and Germany) it has made considerable progress.
The definitions of the 'working poor' are as diverse as the definitions of low wages. The most detailed and precise definition is that of France's National Institute of Statistics and Economic Studies (Institut national de la statistique et des études économiques, INSEE) which covers workers who: have spent at least six months of the year on the labour market; have have had a job for at least one month during the year; and live in a household, in which the income per 'unit of consumption' is below 50% of the median income. The commonly accepted German definition is an extension of the concept of low pay: when their wages fall below 50% of the national average, workers in full-time employment are considered as 'working poor'. A poverty report drawn up by the Hans Böckler Foundation, the German Confederation of Trade Unions (Deutscher Gewerkschaftsbund, DGB) and the Equal Representation Charitable Association (Paritätischer Wohlfahrtsverband) applies a threshold of 40% of average pay for defining 'extreme' poverty, 50% for the 'middle level' of poverty and 75% for 'precarious prosperity'.
In many countries there is no specific definition of 'working poor' and the point of departure for assessing poverty is household income, not each individual’s income. Commonly used definitions include households with at least one income from full or part-time employment, whose incomes are less than half or 60% of the median income (UK), or households with a disposable income after tax below 40%, 50% or 60% of the median or average income (Norway), or households with a low income and insufficient 'social chances' (Austria).
The extent of low pay
The available statistics on the number of low-wage workers across the EU and Norway (both in absolute numbers and as a percentage of employees) show that low pay in is relatively common, although its extent varies between countries, as well as depending on the definition used. Table 1 below sets out the proportion of all employees who are low-paid according to the definitions commonly used in each country, providing data for the most recent available year.
A rigorous comparison of the national data on the incidence of low pay in the various countries is not possible since definitions vary substantially, research is not abundant and the year of reference is different. Therefore we include in the table figures produced on a uniform basis by Eurostat in 2000 (Statistics in focus, Population and social conditions No.11/2000, Low wage employees in EU countries). These figures define low-wage employees as those whose monthly wage is less than 60% of the median national salary, and are based on the 1996 European Community Household Panel (ECHP) survey. No relevant data are available from national sources for Belgium, Denmark, Finland, Luxembourg, the Netherlands or Sweden, and no data from Eurostat for Finland and Sweden (which are thus omitted from the table). Broadly speaking, between 10% and 20% of employees in the countries considered are low paid. Eurostat puts the average for the 13 countries in its research at 15%, or nearly one employee in seven. This figure is as high as one in five in the UK, with Ireland, Germany and Greece also having a high level of low pay. The lowest level is in Portugal, at around one in 17, followed by Denmark and Belgium.
|Country (year)||National statistics||Eurostat statistics*|
|.||Low-wage workers as a % of employees||Definition||Other||.|
|Austria (1993)||13.2%||Under two-thirds of median earnings of full time employees.||Figures cover full-time employees only. 22.8% for women, 7.0% for men, highest level (19.5%) among under-25s.||16%|
|France (2001)||16.6%||All low-wage workers - ie those earning less than two-thirds of the median monthly wage.||79% of all low-wage workers are women. Highest levels among part-time workers and those aged 16-20.||13%|
|(2001)||9.4%||Very low-wage workers - ie those earning less than half the median monthly wage.||80% of very low-wage workers are women. Highest levels among part-time workers and those aged 16-20.|
|Western Germany (1997)||35.5%||Income below 75% of the average national wage of all full-time employees.||Figures cover full-time employees only.||17% (whole of Germany)|
|(1997)||11.5%||Income below 50% of the average national wage of all full-time employees.||Figures cover full-time employees only.|
|Eastern Germany (1997)||35.6%||Income below 75% of the average national wage of all full-time employees.||Figures cover full-time employees only.|
|(1997)||9.5%||Income below 50% of the average national wage of all full-time employees.||Figures cover full-time employees only.|
|Greece (1995)||16.1%||Below two-thirds of median gross monthly wage.||19.8% in services, 14.3% in industry.||17%|
|Ireland (1997)||22.0%||Low-wage workers - ie with earnings below two thirds of median earnings||-||18%|
|(1997)||11.0%||Very low-wage workers - ie with earnings below half median earnings .||-|
|Italy (1998)||18.3%||Less than two-thirds of median earnings for all full-time workers.||Figures cover all employees. 25.9% for women, 13.0% for men. Highest levels among those aged under 30.||10%|
|Norway (1999)||22.0%||Below 85% of the average hourly wage of blue-collar workers in manufacturing.||30% for women, 14% for men.||-|
|Portugal (1998)||11.6%||Below two-thirds of median annual earnings of full-time employees.||-||6%|
|Spain (date unknown)||22.0%||Below 75% of average pay.||-||13%|
|UK (2001)||1.3%||Earning less than national minimum wage.||-||21%|
|(2001)||17.3%||Below half male median earnings (Low Pay Commission definition).||-|
|(2001)||27.6%||Below half male median earnings (Low Pay Unit definition).||-|
* Monthly wage less than 60% of the median national salary.
Sources: for details of the sources of the national statistics, see the national inputs to this comparative study, available on www.eurofound.europa.eu/eiro; Eurostat statistics taken from Statistics in focus, Population and social conditions No.11/2000.
Where such data are available from national sources, it is clear that women are much more likely than men to be low paid. The Eurostat figures indicate that 77% of all low-paid workers are women - a level rising to as high as 86% in Austria, and 81% in the Netherlands and the UK. The lowest proportion of women among low-paid workers is found in Greece (51%) and Italy (60%).
Information on the source, coverage and quality of the low-wage statistics in the various countries examined is provided in the annex at the end of this study.
Trends in low pay
The lack of relevant statistics in some countries make it difficult to assess trends in the incidence of low pay in these cases. Furthermore the data which are available are often not very up to date. Where information is available, the picture varies across Europe. It seems that the incidence of low pay has increased - though not always consistently - over the past 10 or 20 years in countries such as Austria, Belgium, France, Germany, Ireland, Italy and Spain. For example:
- in France, the proportion of low-wage workers rose from 11.4% in 1983 to 13.4% in 1989, then from 15.7% in 1990 to 16.6% by 2001. Although irregular, this increase continued practically uninterrupted until 1995, when the level of low-wage workers reached its high point for the decade (18.4%), before falling in 1999. This rise was even steeper for very low-wage workers, a group which grew rapidly between 1983 and 1998, rising from 5% to 10.6%, before dropping to a little over 9% in 2001. In other words, there has been not only an increase in the number of low-wage workers, but, within that group, a rise in the proportion of the least-well paid;
- in Germany, there was a significant rise in the percentage of workers earning less than 75% of the average national wage of all full-time employees between 1975 and 1986. The incidence of low pay has continued since, but decelerated substantially between 1986 and 1997; and
- in Italy, the incidence of low pay fell sharply at the beginning of the 1980s before rising slightly over the decade and then falling back again in 1989. It grew to 15.7% in 1993, then dropped to 13.7% in 1995 and surged ahead to 18.3% in 1998.
However, not all countries have experienced an increased incidence of low pay. The situation seems stable in Denmark, while the proportion of low-paid workers appears to have dropped in Norway (particularly among female wage earners, but also to some extent among men) and Portugal (where the incidence of low pay peaked at 15.5% in 1993 and has fallen since, standing at 11.6% in 1998). In the UK, the number of low paid workers increased substantially during the 1980s and for most of the 1990s. However, as a result of the introduction of the national minimum wage in 1999, the number of very low paid workers has declined significantly.
How much do low-paid workers earn?
The various definitions of low pay as a proportion of national median/average wages do not indicate what the actual cash amount of low pay might be in each country. National statistics and research rarely allow for an exact calculation of how much low-paid workers earn, but there are a number of measures which may allow for a rough estimation, such as minimum wage rates or the application to average earnings figures of the various thresholds for defining low pay. For example:
- in 2000, some 30% of all Austrian employees (including atypical/part time workers) earned less than EUR 12,831 gross per year. Among all employees fully covered by the social security system, in 1997 the monthly median gross earnings (inclusive of additional payments) amounted to EUR 1,330 (ATS 18,300) for women, and EUR 1,840 (ATS 25,300) for men, while the gross monthly earnings of the lowest paid 20% of female workers came to a median value of EUR 530 (ATS 7,300), with the equivalent figure for male workers standing at EUR 930 (ATS 12,800);
- Belgium has a 'guaranteed monthly average minimum income' (Revenu minimum mensuel moyen garanti, RMMMG) - the minimum wage that private sector employers must guarantee full-time workers during an average month. The RMMMG stands at EUR 1,163.02 per month in 2002;
- in Denmark, assuming that the low-wage threshold is 60% of the median hourly wage, this gives an hourly cash value of DKK 103 (EUR 13.85). This compares with an average collectively agreed minimum hourly wage of DKK 85 (EUR 11.40) and a highest unemployment allowance of DKK 82 (EUR 11) per hour;
- the average gross monthly income of full-time employees, including dependent managers, in 2001 stood at EUR 2,706 in western Germany and EUR 1,817 in eastern Germany. Taking 75% of this amount as the threshold for low pay gives figures of EUR 2,029 for western Germany and EUR 1,363 for eastern Germany. The 50% threshold for very low pay translates as EUR 1,329 for western Germany and EUR 909 for eastern Germany. A calculation based on the distribution of wages in Germany indicates that the average wage of low-wage workers (according to the German definition) is 55% of the average wage, or EUR 1,485 per month;
- in Greece, the average low wage in 1995 was calculated as EUR 367 per month and EUR 2.8 per hour;
- in Ireland, low pay at the threshold of two-thirds of average earnings means pay of around EUR 200 per week;
- in the Netherlands, the best approximation for low wages is the statutory minimum wage. The gross annual minimum wage for those aged 23 years and above stood at EUR 13,975 in 1999 (corresponding to approximately 60% of the average wage). Some 1.86 million employees earned less than this amount over a full year, representing 30% of all employees;
- in Portugal, the average low wage in 1998 was calculated at EUR 325 per month; and
- in Sweden, the lowest wages are found in the local government sector (schools, child daycare centres, care of the elderly, hospitals etc), where the average monthly wage stood at EUR 1,545 in 2001 - around 70% of the average wage for all employees, which stood at EUR 2,164 (including blue-collar, white-collar and professional professional workers).
Do low-paid jobs lead to better-paid jobs?
It is often argued that low-paid jobs lead to better-paid jobs and they serve as a useful entry point to employment for those with few qualifications or little work experience. In some cases, this seems to be confirmed by national evidence. For example, in Denmark, a study on 'individual wage mobility' over 1980-90 found that people in low-wage jobs move relatively quickly from the lowest to the upper deciles of the wage distribution. Less than 20% remain in the lowest deciles for more than four years. The conclusion was that, first, it is important to be employed even at a low wage in order to obtain better employment possibilities later, and second, the risk of remaining in low-wage jobs is small.
However, the evidence from a number of other countries does not support the idea that low-paid work serves as an entry point to upper segments of the labour market. For example:
- in France, the rise in the number of low-wage workers over recent years has gone together with an increase in the risk of remaining a low-wage worker, at least in the short term. Furthermore, movement between unemployment and low-paid employment has become more frequent. The percentage of people who have always been either in low-wage work or unemployed almost doubled in the 1983-2000 period, rising from 7.3% over 1983-5, to 14.1% in 1998-2000;
- in the UK, although a number of low paid jobs are associated with young workers, including students who move on to better paid jobs and others who improve their position through job tenure and experience, research suggests that those who are in low-paid work have a higher propensity to be unemployed and to remain in low paid work. A study over the period 1991-4 found that 56% of men moving out of unemployment moved into jobs with wages in the bottom 25% of the earnings distribution. Of the men in this bottom 25% in 1991, half were still there by 1994, 13% were out of work and only 6% were in the top half of the earnings distribution. For women who were in the bottom 25% in 1991, 44% were still in the bottom quarter three years later, 22% were out of work and 10% were in the top half of the distribution. Half of all men who were persistently in the bottom 25% were poorly educated and/or young;
- research in Portugal found that 40% of low-wage workers in 1996 had been low-wage workers in 1991, while 60% of those who were low-wage workers in 1991 had ceased to earn low wages in 1996;
- in Austria, there is little evidence that low-paid jobs lead to better-paid jobs to a significant extent, with low pay correlated with low qualifications and the interruption of careers. A longitudinal analysis of earnings mobility from workers' first entry point into the labour market in 1977 up to 1996 found that especially female low-wage workers (in the lowest 20% of the wage distribution) were unable to improve their position significantly; and
- according to an analysis covering the 1975-93 period, the status of low-paid workers in Italy seems to be relatively persistent, with 50% of the individuals who were in the lowest decile of the wage distribution at the beginning of the period still in the same decile at the end of the period. This percentage rises to 60%, if the bottom two deciles are considered. Another study found that a significant proportion of workers in the lowest decile exit from it over time. However, a substantial proportion of these departures are accounted for by low-wage workers become unemployed (a fact suggesting that there might be a low-paid employment/unemployment 'trap').
Factors affecting low wages
The overall factors contributing to low-paid work are, first, part-time work, and second, low hourly wage rates. According to the Eurostat survey referred to above, the reasons for low pay are primarily the 'part-time effect' and the 'low-remuneration effect', with the former accounting for 43% of low-paid work in the 13 EU countries examined and the latter for 37%. A combination of the two is the reason for 11% of low-paid work, with only 8% accounted for by neither.
However, behind these factors are issues related to the individual characteristics of low-paid workers, the sectors and establishments in which they work and they types of job they do. Low-paid workers have similar characteristics across all the countries considered, despite the large differences in the overall incidence of low pay. The most common factors reported as influencing low pay are: the size of establishments (which can often be relevant to the presence and strength of trade unions); sector/industry; occupation; type of ownership of the establishment (public, private, 'third' sector), the marital status, gender, age, education and skills of the worker; and the type of contract and type of employment. For example, the incidence of low hourly wage rates tends to decrease with age and differs among sexes. In sectoral terms, low-wage sectors are often to be found in private service industries with a predominately female workforce, in local government and work of 'public utility' (childcare, care of elderly and disabled people, cleaning services etc). Low pay tends particularly to affect women, young people, unskilled blue-collar jobs, people on fixed-term contracts, trainees, people with a short period of service with their employer, working pensioners, students, widowed, divorced or single persons, and immigrants.
The degree to which each of the above factors affects the extent and the persistence of low wages varies across countries, as indicated in table 2 below (no information is available for Luxembourg).
|Austria||The sectors paying the lowest wages are agriculture, hotels and catering, leather production, public and private services and the textiles industries. Small establishment size plays a crucial role (due to lower trade union presence and bargaining power). Owing to low productivity and the weak position of the unions, many branches of the private services sector provide low wages even for full-time workers. The most important individual factors affecting low wages are gender (with women particularly likely to be low paid), marital status (with low pay among lone mothers), a low level of education and skills, and nationality (with many immigrants low paid).|
|Belgium||Low pay is particularly prevalent among low-skilled and women and in the unproductive not-for-profit sector.|
|Denmark||Sector/industry,,occupation, gender (with women more likely to be low paid), and a low level of education and skills are reported as factors favouring low pay. Low-wage sectors and occupations include the retail trade, cleaning services, unskilled childcare work (in the public sector), restaurants and hotels (or tourism as a whole), assembly work in the electronics industry (mostly involving female workers) and some unskilled jobs in construction.|
|Finland||Low-paid work is concentrated in the service sector and public sector and particularly affects employees in temporary work.|
|France||The risk of low-wage work is related, first of all, to the characteristics of the jobs occupied. Part-time jobs account for 77% of low-wage workers, while fixed-term jobs and those created under employment schemes are over-represented. Civil servants are less likely to be low-wage workers, while commerce and service sector employees and unskilled workers are more exposed to the risk of low wages. In terms of personal profiles, the vast majority of low-wage workers are women, young people, and those with low qualifications.|
|Germany||The factors of part-time work and a low hourly wage rate affect almost equally the probability of earning low wages. Among factors influencing a low hourly wage rate, most important seem to be jobs in services, gender (with women making up the majority of low-paid workers) and a low educational level. A particular regional factor in Germany is the division between western and eastern Germany, with the latter arguably nearly a low-wage region.|
|Greece||The factors correlated with low wages are age (with younger workers more affected), length of service, low educational level, low union density in the establishment, gender (more women are low paid), marital status (single workers are over-represented among low-wage workers), fixed-term contracts, part-time work, low-paid occupations and sectors, low hierarchical position, no access to overtime and shiftwork, and small establishment size.|
|Ireland||The characteristics distinguishing low-paid employees from other employees are part-time employment and being under the age of 25. There is no difference in risk between men and women under the age of 25, but women above that age face a substantially higher risk of being low paid than men – with married women being particularly vulnerable to low pay. The risk of low pay is also closely linked to levels of educational attainment and skill, although this varies by age. There is strong evidence of a widening gap between the more and less skilled and educated. The risk of being low paid also varies across industries and occupations, and is particularly high in retail and personal services.|
|Italy||Research indicates that the probability of being low-paid is greater for women, people in manual jobs, young people, less-educated and smaller firms workers. The retail and personal services sectors have a higher than average level of low-wage work. Finally, in regional terms, workers in the South of Italy have a higher probability of being low-paid than those in the Centre or the North of the country.|
|Netherlands||Key factors influencing low pay are the very high incidence of part-time work, the high incidence of flexible contracts, and the low statutory minimum wages for young employees under 23 years of age.|
|Norway||Low-wage workers tend to be women and young people. Wage levels (on average) are lower among certain low-skill occupational groups or sectors/industries- notably hotels and restaurants, the textiles industry and some groups of employees in the municipal sector.|
|Portugal||According to research conducted in 1996, the main factors influencing low wages are gender (with women twice as likely as men to have low pay), age (with workers under 25 particularly affected), a low level of education and qualifications, length of service with the employers (with employees with under five years' service more likely to be low paid), and small establishment size.|
|Spain||Low pay is correlated mostly with gender (with women earning less pay than men for same job, facing occupational discrimination and over-represented in low-pay sectors and temporary jobs), temporary employment (which affected 34.1% of wage earners in the private sector and 22.1% in the public sector in the fourth quarter of 2001), part-time employment (which makes up 8% of total employment), temporary agency work, and training contracts (which are common particularly among people without tertiary education).|
|Sweden||Low-paid workers are typically female employees in the local government sector (often part-time workers) and employees in the hotels and restaurants sector.|
|UK||Research indicates that in 1997, the following were disproportionately affected by low pay: women, young people, ethnic minorities, long-term disabled people, part-time workers, lone parents, seasonal workers, casual workers, temporary agency workers and home workers. The job characteristics associated with low pay were manual work, work in the private/voluntary sector and retail and wholesale trade, work in establishments with no trade union recognition, and work in small establishments (under 25 employees). Half of all low paid work was in retail, business services (contract cleaning and security), social care, hairdressing, hospitality, agriculture and clothing and footwear. Another piece of research, conducted in 1998 and covering workplaces with 25 employees, found that across the private sector 13% of workplaces had more than one quarter of workers who were low paid (defined as earning less than EUR 5.53 per hour). This figure was higher where more women were employed and where part-time workers were in the majority. Another study based on the same data found that the low paid were more likely to be in private sector establishments in competitive product markets, with limited collective industrial relations institutions.|
Table 3 below provides possible reasons why some of the most commonly cited factors - in terms of both the employee's personal characteristics and the nature of their job and its industrial relations context - might contribute to low pay.
|Age of the worker||Younger workers have less accumulated work experience.|
|Educational level||Workers with a low level of education and qualifications have a lower level of knowledge, skills and productivity.|
|Length of service with same employer||Workers with shorter service have less recent relevant work experience.|
|Marital status||Marital status may affect workers' social skills.|
|Gender and nationality||Women and foreign workers may face sex and racial discrimination.|
|Characteristics of job and industrial relations context||.|
|Overtime and shiftwork||Lack of access to these forms of work will reduce a worker's ability to increase earnings.|
|Hierarchical position||Employees who do not supervise other workers will not receive the resulting additional pay.|
|Presence and strength of trade unions in workplace and relevant bargaining level (sector level, enterprise level)||These factors may affect the balance of power between workers and employers at the workplace, both individually and collectively.|
|Fixed-term contracts and seasonal work|
|Size of establishment|
|Local government, 'third sector', retail and wholesale, hotels and restaurants, cleaning, security services, clothing and footwear, personal services||Jobs in these sectors may have low productivity.|
|Part-time work||The lower hours of work result in lower overall pay.|
|Jobs created under public employment schemes, temporary agency work, home work||This type of employment may have a lower than usual hourly wage rate, and may have low productivity.|
Low wages and low productivity
On the question of whether low-wage work is concentrated in low-productivity sectors of activity, and whether there are low-wage workers in more high-productivity and 'knowledge-based' activities, most of the national evidence confirms that low pay is generally associated with low productivity.
It is reported from countries such as Austria, Finland, Greece, Ireland, Italy, the Netherlands, Portugal (where low productivity is a feature of the whole economy) Spain and the UK that low wages are normally associated with low-productivity activities. However, in many cases - such as Denmark, Germany and the UK - much low-paid work is concentrated in the service sector, where productivity is difficult to measure. There are exceptions to the low-wage/low-productivity rule. In the cases of Denmark and Germany, it is reported that there is no strict correspondence between low wages and low productivity.
It appears that low-paid workers are not entirely absent from 'knowledge-based' activities, and their presence is reported from countries such as Denmark (where many students and young people may work part time in these industries), Finland, Portugal (eg young researchers) the UK (eg warehousing and shipping goods, customer services and call centres).
Low wages and the 'working poor'
The concept of 'working poor'- ie those who are in employment but remain below a defined poverty threshold - is not the same as low-paid work, though the latter may of course contribute to the former. For example, low-paid workers may live in households with other sources of income, such as the earnings of other household members, or social benefits, and thus avoid poverty. National information is limited concerning the number of 'working poor' in the countries examined and the relationship between low wages and poverty among working people. However, the abovementioned Eurostat study provides some standardised information on these issues for 13 EU Member States (excluding Finland and Sweden).
The Eurostat statistics, which refer to 1996, provide indications as to the number of 'working poor' in the 13 EU Member States - see table 4 below (there is no gender breakdown available). The table shows the proportion of all paid employees (working at least 15 hours per week) who are 'poor'- that is, who have an 'equivalised income' (ie the total household income divided by the number of adult equivalents in the household) which is below a low-income threshold of 60% of the national median equivalised income. This gives an indication of the proportion of all employees who are 'working poor', The table also shows the proportion of low-wage employees (those whose monthly wage is less than 60% of the median national salary) who are poor, and the proportion of all poor employees who are low-wage employees.
|Country||Poor employees as % of all employees||Poor employees as % of all low-wage employees||Low-wage employees as % of all poor employees|
Note: see main text for definitions of terms.
Source: Eurostat, Statistics in focus, Population and social conditions No.11/2000.
Table 4 indicates that 8% of employees in the EU countries considered are 'poor', with the highest levels of such 'working poor' in Germany and Italy, and the lowest in Denmark and Portugal. As mentioned above, national statistics on the working poor are rare and based on varying definitions, but relevant information is available for a few countries:
- a 1998 study on poverty in Austria found a high proportion of 'working poor' among poor people. Excluding pensioners, almost 50% of poor people were wage earners and 10% were self-employed. Furthermore, it is estimated that 25% of unskilled workers, 14% of manual workers and 8% of salaried employees are in danger of 'pauperisation' in the event of an accumulation of risk factors;
- in France, the number of poor workers was put at between 1.3 million and 3.1 million, depending on the definition used, in the mid-1990s. There are no statistics plotting changes in the number of working poor over time, but certain indices suggest that their number of these workers has probably grown, especially during the first half of the 1990s; and
- in Norway, research indicates that approximately 2% of those working full-time and 4% of those working part-time (up to 35 hours per week) are in 'poor households' (the poverty line is set at 50% of median disposable income). If these estimates were based on individuals rather than households, the proportion of working poor would be lower.
On the relationship between low pay and the working poor, the Eurostat figures in table 4 above indicate that low-wage employees are considerably more likely than the total population of employees to be poor - some 20% of low-wage employees are considered as poor (compared with a general rate of 8%). Low-wage employees are most likely to be poor in Greece, Italy and Germany and least likely to be poor in Denmark and Ireland. However, 80% of low-paid workers are not poor. Furthermore, low-wage employees make up only 37% of all poor employees across the EU - implying that 63% of poor employees are employees who are not low paid. Low-wage employees make up the highest proportion of all poor employees in Greece and the UK and the lowest proportion in Denmark, Portugal and Denmark, and there are considerable difference between the Member States in this area.
It is clear that there is not necessarily a direct link between low pay and the working poor - many low-paid workers are not poor and vice versa. This borne out by the available national data on this issue. For example:
- a majority of French low-wage workers - 60%-80% depending on the definitions used - are not poor. Given that the incidence of poverty among employees is much lower than that of low wages, even if all poor workers were low-wage workers, the poverty rate among all low-wage workers would still be much less than 40%;
- Irish research has found that the overlap between low pay and poverty is weaker than often assumed. Only a small minority of the low paid are in households that would generally be identified as being poor. Many households characterised as poor do not have any member working at all. This is a reflection of the tendency towards a polarisation between work-rich and 'work-poor' households;
- low pay does not seem to be the main cause of poverty in Spain, where the key role is played by inactivity and unemployment, given low retirement pensions and unemployment benefits. Only 50% of unemployed people receive any type of benefit and only 50% of recipients receive an income equivalent only to a third of the average wage. For these reasons there is little relation between poverty and low pay, with the sole exception of lone parent households where a sole source of wage income and low pay coincide; and
- a 1997 study found that during the 1970s and 1980s, only about 3% to 4% of low-paid individuals in the UK lived in 'household poverty' (defined as less than 50% of average household income). By the early 1990s, the proportion had increased to 13%. These tended to be low-paid lone mothers and couples including only one earner, and represented around 600,000 working poor and nearly 3% of all households. The study estimated that around 87% of the low-paid were not working poor;
As to the factors which prevent low-paid workers from falling into the category of the working poor, in most countries the most important are: a second (or more) wage earned by a partner or other household member; and social transfers, notably housing benefits or childcare-related allowances, plus minimum income schemes in some countries. In some southern countries an important factor is a long tradition of inclusive families, with strong ties of solidarity between members (as in Spain and Greece), along with additional property income (as in Greece) or income from farming and similar additional activity in rural areas (as in Portugal).
Minimum wages and collective bargaining
The above evidence indicates that in most of the countries considered, and despite the large differences across countries and sectors, the proportion of low-wage employees is relatively high - with an EU average of 15% according to the Eurostat figures. In this context, tools and policies to address low pay are widely regarded as necessary in both the public and the private sector. Among such instruments, commentators are agreed that the legislative and institutional setting for pay determination is of major importance. In particular, low pay researchers argue that minimum wage regimes reduce the number of low-paid workers, suggesting that the incidence of low pay can be expected to be lower in countries with a minimum wage.
Minimum wage systems
Minimum wages of one form or another exist in all countries examined. The systems vary considerably, and can broadly be classified in two groups.
In the first group of nine countries, there is a national minimum wage with legal force. In France, Ireland, Luxembourg, the Netherlands, Portugal, Spain and the UK a statutory national minimum wage is fixed at an hourly, weekly or monthly rate by the government, in many cases in consultation with the social partners, and this minimum is enforced by law. In Belgium, the RMMMG monthly minimum wage for the private sector is set by an intersectoral collective agreement concluded in the National Labour Council (Conseil National du Travail/Nationale Arbeidsraad, CNT/NAR), which is legally binding. In Greece, minimum wage and salary rates for blue- and white-collar workers respectively are set by the social partners in the National General Collective Agreement (EGSSE), which covers all workers in the private sector (GR0204109F). In all of these countries, the legal minima may be supplemented by minimum rates set in collective agreements - though the level and coverage of such agreements varies, as does the relationship of the minima they set to the national minimum rate.
In the second group of seven countries - Austria, Denmark, Finland, Germany, Italy, Norway and Sweden - there is no national minimum wage, but sectoral collective agreements implicitly or explicitly set minimum pay rates for the employees in their area of coverage. This thus results in a variety of minimum wage rates for different sectors. The impact of these minimum rates varies, so that this group of countries can be further divided, as follows.
In Austria, Finland, Germany and Italy, the coverage of collectively agreed minimum rates is enhanced by various means. Finland and Germany have mechanisms for the extension of sectoral agreements to employers and employees in the sector which are not members of the signatory organisations. For example, the German system of extension through agreements being declared 'generally binding' (Allgemeinverbindlichkeitserklärung, AVE) on non-members seeks to protect employees from 'wage dumping' and employers from unfair competition. In 2002, some 516 of about 55,000 registered collective agreements had been extended (down from 534 in 2001), and the number of extended agreements on pay stood at 53 in 2001 (down from 61 in the previous year). In Austria, a high level of coverage of sectoral agreements and their minimum wage rates is guaranteed by the fact that all employers are obliged to be members of the Chambers of the Economy (Wirtschaftskammer) which conclude collective agreements (AT9912207F). In Italy, case law has established that collectively agreed sectoral minimum pay rates must be applied by employers that are not members of signatory organisations. Denmark, Norway and Sweden, on the other hand, have no system of extending collective agreements, or any similar mechanism.
The relationship between pay minima agreed at sector level and actual pay also differs between countries in this group. In countries such as Denmark, Norway and Sweden, many sectoral collective agreements provide specifically for additional local-level bargaining on top of the central increases, with only relatively few sectoral agreements (as in some cases in Denmark and Norway) setting actual pay rates. Some agreements in the Norwegian private sector include 'guarantee clauses', providing for pay adjustments aimed at at raising the wage level among the groups with the lowest pay or companies with the lowest average wage level. Sectoral pay rates are also often improved on at lower level in countries such as Austria, France and Germany.
Impact on low pay
Overall, the national evidence seems to confirm the proposition that minimum wage regimes reduce the number of low-paid workers. Obviously, a minimum wage provides a pay 'floor' for those workers covered by it, while it may also have the effect of pushing up other wage rates which are above the minimum level (as reported, for example, from Luxembourg), or have a 'knock-on' effect by influencing pay rates in areas not covered by the minimum wage (as, for example, in Denmark). However, there are a number of important factors which influence the impact of minimum wages on low pay, considered below. It should, though, be noted that the existence in all the countries examined of a statutory or collectively-agreed minimum has not - as seen above - effectively hindered the persistence of low pay among large groups of employees, particularly women, part-time, short-term and seasonal workers, immigrants and students.
The scope of minimum wages is clearly a key issue. In those countries with a national minimum wage, this tends broadly to cover all categories of employee, though in some cases, such as Belgium and Greece, it covers only the private sector. However, in many countries, certain categories of workers are either excluded from coverage by the minimum wage or receive a lower rate. Younger workers receive a lower minimum wage rate (sometimes progressing in stages with age) in Belgium (where full minimum wage entitlement starts from the age of 22), Ireland (18), Luxembourg (18), the Netherlands (23) and the UK (22). Greece has a separate rate for blue- and white-collar workers, while Luxembourg distinguishes between skilled and unskilled workers. A length-of-service requirement is imposed for entitlement to the full rate in Belgium and Ireland. In Ireland, Portugal and the UK, workers undergoing training receive a lower rate. In Luxembourg, the Netherlands and Portugal, a lower rate may be paid to workers with a reduced capacity to work. In Portugal, there is a separate, lower rate for domestic workers. In the UK, a relatively large number of categories are excluded entirely, including workers under the age of 18, apprentices under 19, apprentices under 26 in the first 12 months of apprenticeship, some trainees, students doing work experience as part of a higher education course, company directors and people living and working within a family. These various lower rates and exclusions may contribute to the concentration of low-wage workers among groups such as young people and unskilled workers.
In countries where sectoral collective agreements play a sole or important role in minimum wage setting, the impact is conditioned by the coverage of these agreements. Unless sectoral agreements cover all sectors, and all employees within these sectors, then their minimum wage protection is limited. Even in countries with a high rate of coverage by sectoral bargaining, there are sectors which have no agreement - often in private service sectors. Furthermore, bargaining coverage tends to be lower in small-sized enterprises. The widely-reported concentration of low pay in services and smaller firms may be attributable, at least in part, to these factors. For example, it is reported from Germany - where sectoral agreements play the key role in minimum pay-setting - that the main causes for low pay include the existence of sectors which are not covered by a collective agreement, along with situations where employers breach the pay terms of collective agreements, or leave the relevant employers' association and are thus no longer covered by a sectoral agreement.
Minimum pay levels
Another vital factor in the effectiveness of minimum wages as a means to improve the pay situation of low-paid workers is obviously the level at which the minimum wage is set: if the level of the minimum wage is too low, the incidence of low pay will not be dramatically affected. In absolute cash terms, there are very large differences in minimum wage rates between countries. As indicated by the Eurostat data in table 5 below, which covers countries with a national minimum wage system, the level of the minimum wage in Luxembourg and Netherlands is three times higher than that in Portugal. Monthly minimum rates stand at over EUR 1,000 in Belgium, France, Ireland, Luxembourg, the Netherlands and the UK, but around EUR 500 or below in Greece, Portugal and Spain. Eurostat attempts to give an idea of the real level of the minimum wage in national terms by comparing it with average earnings in manufacturing in each country. This suggests that, in relative terms, the national minimum wage is highest in Portugal (57%) and lowest in Spain (34%), with a rate of over 40% of average earnings also found in France, the Netherlands, Luxembourg and Greece, and a rate of below 40% in Belgium and the UK.
|Country||Monthly minimum wage (EUR) (2002)||Minimum wage as % of average earnings in manufacturing (1999)|
The Eurostat data do not cover countries with a collective bargaining-based minimum wage system, but the national evidence available for a few such countries suggests that these tend to set relatively high minimum rates, compared with many national/statutory systems. For example, the influential hourly minimum wage in the Danish services sector is just over half of the median hourly wage, while German research indicates that the minimum wage rates set by industrial collective agreements stood at 58% of the average wage in 1993. However, it should be noted that agreed minimum rates can vary considerably between sectors, as for example in Belgium.
A relatively high level of minimum wage is likely to reduce the number of low paid workers considerably, especially where combined with combined with a small variance in wage distribution (as, for example in Denmark). However, where the minimum wage is set at a low level it is likely to have only a negligible affect on low-paid workers, most of whom will be paid above this rate. A notable example is Spain, where the minimum wage is so low - in terms of both absolute and relative values, it is one of the EU's lowest (ES0201249N) - that it does not act as a basic reference point for pay-setting on the labour market (except for training contracts, for which pay is based on this minimum rate).
An indication of the direct impact of minimum wages is the proportion of workers who are paid at this rate. According to Eurostat, in 2000-1 only around 2% of employees in Ireland, the Netherlands, Spain and the UK earned the national minimum wage, compared with over 6% in Portugal, nearly 14% in France and 15.5% in Luxembourg. This would tend to bear out the idea that higher national minimum wage rates have a greater real impact.
In some of the countries considered, it is argued in some quarters that national minimum wages are not set at a level high enough to ameliorate the worst effects of low pay and poverty. In this context it is unsurprising that some trade unions in countries such as France, Greece, Ireland, Spain and the UK are demanding substantial rises in the minimum wage (see below under 'Social partner actions and views').
In collective bargaining-based systems, the proportion of workers who are paid at this rate is rather more difficult to assess, but the evidence that is available indicates variations among sectors and employee groups. The minima set by sectoral agreements are - as mentioned above - often exceeded in practice, as a result of lower-level bargaining or other arrangements. In Austria, for example, the importance of minimum pay rates varies across sectors - in some sectors (such as catering) the minimum wages are the actual rates paid, in other sectors the majority of workers receive rates above this level. In Denmark, collectively agreed minimum rates are usually increased by local-level bargaining, and minimum rates are the actual pay rates in only a small minority of cases, such as in some service sectors (eg retail and hotels and restaurants), and for specific groups such as seasonal or short-term workers.
There is no evidence that the setting of sectoral minimum wage rates is particularly prevalent in the main low-pay sectors - generally retail, hotels, restaurants and catering, textiles manufacturing, cleaning, agriculture and building - and no country has a system applying a minimum wage solely to these sectors (though Ireland and the UK had such a selective minimum wage system before recently introducing national minimum wages). However, in Germany, for example, the practice of extending collective agreements is used particularly in low-wage branches such as building, retail, catering, textiles and cleaning.
Relationship between national and sectoral rates
In countries with both a national minimum wage and a high coverage of sectoral collective agreements setting minimum pay rates, the impact of the former depends very much on the level of the latter. In Belgium, for example, the minima set by sectoral bargaining exceed the national minimum wage by around 15%-30%. By contrast, in France, the lowest minimum rates set by collective agreement have, in the majority of cases, been overtaken by the SMIC national minimum wage, rendering them ineffective.
There generally seems to be little evidence or debate on whether national/statutory minimum wage systems or sectoral/agreement-based systems are more effective in dealing with low pay. There are, however, some parties who argue that the former system is preferable. For example, some German commentators state that the absence of a statutory national minimum wage in that country has promoted low pay, as collective agreements do not apply to all employees and do not guarantee wages at an appropriate minimum level.
Evidence from Ireland and the UK
The impact of a statutory national minimum wage on low-paid workers, compared with other factors, is hard to detect precisely in countries with a long-standing minimum wage. However, the very recent introduction of a national minimum wage in Ireland and the UK makes analysis of this point considerably easier in these countries.
In Ireland, the national minimum wage introduced in April 2000 (IE0107170F) has undoubtedly been important for raising income levels among low paid workers (and facilitating a small, albeit important, reduction in the gender wage gap). Research published in 2001 indicates that around 5% of employees had their pay increased as a direct result of the introduction of the minimum wage (IE0110103N) - a lower figure than was initially envisaged. It appears that tight labour market conditions and labour shortages were a more significant factor behind wage increases in many low-paying sectors, as unemployment dropped sharply from the late 1990s. The study found that as many as 81% of employers claimed they would have increased wages to attract or retain labour, regardless of the minimum wage.
The Irish evidence also underlines the concentration of low pay in certain sectors and smaller firms, and the effect that a national minimum wage can have. The percentage of employees who had a wage increase as a direct result of the minimum wage was highest in retail (11.6%), followed by hotels, restaurants and bars (8.9%), and textile manufacturing (7%) - all traditional low-paying sectors. The effect was less in other sectors - wholesale (5.1%), non-textile manufacturing (2.7%), banking and finance (2.4%) and building (0.8%). Only 2.5% of employees in firms with over 100 workers were directly affected by the minimum wage, while in firms with between four and 100 workers, 7% to 8% of workers were directly affected.
Evidence from UK suggests that the introduction of the national minimum wage in April 1999 has had a significant impact on very low-paid workers (UK9904196F). In 1998, 1,520,000 employees were paid less than the initial rate of the minimum wage, while in 2001 only 320,000 were paid less than increased rate applying in that year - a reduction from 6.4% of the workforce to 1.3%.
Considering the high concentration of low pay among part-time workers across the EU and Norway, it might be expected that part-time work would be prominent in the debate on low pay and poverty. However, it appears that all countries covered by this study, except Greece, there are no special legislative or collectively-agreed provisions that focus on minimum pay for part-time workers, providing for example for an increased rate. In Greece, recent legislation (Law 2874/2000) increased pay for part-time employment by 7.5% for working days of one to three hours (GR0012192F). In all other cases, part-time workers have the same minimum wage rates as full-time workers proportional to their reduced working hours. There also appears to be very little debate on the issue, with two exceptions:
- France's SMIC minimum wage an hourly rate and thus enables limits to be put on the growth of low-wage work measured in hourly rates. However, with the strong growth in part-time work since the early 1990s, this hourly wage guarantee is reportedly becoming decreasingly effective in limiting the growth in the number of low-wage workers if looked at in terms of the income it secures for employees (on a monthly or annual basis). Accordingly, trade unions call for the SMIC to be raised, along with more restrictive regulations on the use of part-time work; and
- the Greek General Confederation of Labour (GSEE) is proposing that pay for part-time employment should be 25% higher than the wage corresponding to the reduced duration of work, and under no circumstances should it be less than half the wage corresponding to regular employment.
Decentralisation of bargaining
A trend common to a number of European countries in recent years has been the decentralisation of collective bargaining, generally from sector to company level (though sometimes from intersectoral to sectoral level). On the question of whether the decentralisation of wage bargaining is likely to favour the incidence of low pay, information is available from only a few countries, and indicates that there is no pronounced common trend in this area.
Among countries which have witnessed decentralisation of pay bargaining, the effect on low pay is unclear. In France, the influence of collective bargaining on low-wage workers has generally been weak, as evidenced by the fact that the majority of sector-level minimum wages are lower than the SMIC national minimum wage (see above). However, the decentralisation of pay bargaining has had an impact on the overall pattern of pay levels and has probably contributed to the growth in the number of low-wage workers. In the UK, there has been widespread decentralisation of wage bargaining , and it has been associated with an increase in wage dispersion and wage settlements, but it is uncertain as to how far this has increased the levels of low pay. However, it is reported from Greece that the decentralisation of pay-setting, as reinforced by recent legislation, has created the basis for the deregulation of the contents of collective agreements, with adverse consequences for the pay of a large proportion of wage earners.
Among countries with a more centralised system of wage bargaining, this is seen in some cases as a factor in preventing low pay. For example, it is argued that, to a certain degree, Norway's centralised collective bargaining system is a contributory factor to the relatively small pay differentials that characterise the Norwegian labour market. In Ireland, there is evidence that a 'free-for-all' system of decentralised bargaining, rather than the country's current national pay-setting model, could benefit workers in trade unions with strong bargaining power/industrial muscle, but not so much workers with weaker bargaining power.
'Wage floor' debate
The evidence above indicates that the various minimum wage arrangements across the EU and Norway generally provide a lower threshold for downward pressure on wages, protecting mainly the incomes of low-paid workers who would otherwise receive wages below this level. However, it is often argued - for example, by the OECD - that such minimum 'wage floors' raise the cost of hiring low-skilled workers and those with little working experience, and if set too high relative to the average wage they tend to reduce employment prospects, especially among young and low-productivity workers.
National evidence on this issue is quite sparse. In some countries, such as Austria, Denmark, Norway, Greece, Portugal and Spain, the existence of wage floors is not currently a major issue of debate. Where there is any national information, this tends to oppose the argument that wage floors set by minimum wages have a negative effect on employment. Rather, the evidence suggests neither a positive nor a negative correlation between the level of the minimum wage and the employment of low-paid workers. However, in countries where this question might be more readily assessed because of the recent introduction of a national minimum wage - Ireland and the UK - the effect is unclear because the minimum wage has been introduced in a period of strong growth in the economy and employment. Examples of national information on this issue include the following:
- in France, it is reported that a negative impact of the minimum wage on employment has never been documented on the basis of sound empirical studies. All those studies which posit a negative outcome have, it is claimed, been based on theoretical models whose empirical validity remains unproven;
- a German report on employment creation among low-skilled people indicates that there is no evidence that wage reductions or the creation of new low-wage groups will automatically create substantial employment prospects for these people. Nor is their evidence that a greater wage spread will improve the employment prospects of these groups - for example, the employment share of low-skilled people in Germany is much higher than in countries (such as the USA and UK) with a higher wage spread;
- research examining the effects of the Irish national minimum wage found that none of the firms surveyed said that they had employed fewer people as a result of the minimum wage's introduction, while 5% of firms claimed that they would be employing more workers if the minimum wage had not been introduced. The minimum wage was introduced at a time of strong economic expansion and extremely buoyant employment growth, so its impact on employment was negligible;
- there is reportedly a consensus, even among the more liberal Spanish economists, that the minimum wage for adults has no major influence on employment (doubtless because it is so low). In 1997, the government decided to unify the minimum wage for adults and young people, which involved a rise of 17.7% in the latter. Orthodox economists predicted a fall in youth employment due to this increase. In the three years following the application of this measure, youth employment rose by 22.5%, probably due to a continuous expansion of economic demand and the fact that young people's minimum pay was still low despite the increase; and
- in the UK, the evidence presented to the Low Pay Commission since the introduction of the national minimum wage has generally shown no adverse affects on employment or unemployment. This has been based on both aggregate data and case study research. Some companies have reduced employment, but these have been offset by gains in other companies. The minimum wage was introduced in a period when there was a general expansion in service sector employment, in particularly in those sectors that hire low-paid workers, such as hospitality, retail and business services. Research examining the aggregate impact of the introduction of the national minimum wage has found no adverse employment effects on any groups, including young people.
Trade union bargaining power and membership
Besides minimum wage arrangements, some research suggests that other labour market institutions - and particularly the bargaining power and membership levels of trade unions - can significantly influence the incidence of low pay.
In relation to the bargaining power of unions, what information is available from the countries examined tends to support this proposition. In all eight countries for which some information is available – Belgium, Denmark, Germany, Ireland, the Netherlands, Norway, Spain and the UK- evidence suggests that varying levels of union strength can either reduce or increase the incidence of low pay. In Denmark and Norway the relatively high minimum wage rates and relatively low numbers of working poor are partly seen as a result of a strong bargaining position for unions, while in Belgium and Ireland the present bargaining power of unions is considered as a barrier to low pay.
By contrast, in Germany, the Netherlands, Spain and the UK, economic and political changes, particularly during and since the 1990s, have intensified pressure on trade unions, reducing their capacity to act and promote their demands. In the Netherlands, for example, unions were more or less forced to agree to the introduction of lower wage scales (just above the legal minimum wage) in collective agreements in the mid-1990s, and have not succeeded in their demands to raise the legal minimum wage for employees under 23. Similarly, in Spain the unions have been unsuccessful in their calls for an increase in the low level of the minimum wage to bring it closer to the levels in other European countries (though they have been somewhat more successful in addressing low pay through bargaining). In the UK, there is some limited research that supports the argument that weakening trade unions have allowed wages at the bottom end of the labour market to decline, while unions are still associated with higher wages for their members. Research in the mid-1990s into the impact of falling union membership on the changing wage structure in the UK found that about 20% of the increase in wage inequality was attributable to union decline.
The UK evidence raises the issue of the trade union 'mark-up' or 'wage premium'- the tendency found in a number of countries (such as Ireland and the UK) for trade union members to have higher wages than non-unionised workers. Data on union membership among low-paid workers are available for only four countries – Denmark, the Netherlands, Spain and Sweden - and are too fragmentary and circumstantial to allow a general conclusion to be drawn. However, not surprisingly, two countries with a relatively low proportion of low-wage employees - Denmark and Sweden - have high trade union density in many low pay sectors, while in the Netherlands - a country with a moderate proportion of low-paid workers - trade union density in the lower income brackets (those with a monthly household income lower than EUR 900) is only 5% (national average 28%). Finally, in Spain, the membership of the main trade unions is concentrated in sectors with medium pay levels (such as metalworking, the public sector, chemicals, transport and communications, and banks), while unions find it difficult to organise in the low-pay sectors.
Social partner actions and views
The issue of low pay seems to be relatively high on the agenda of trade unions across the EU and Norway. In principle, in all the 16 countries considered, trade unions share the objective of obtaining an acceptable standard of living for all employees, although the ways in which they seek to do so varies considerably across the different countries, and even across the different trade unions within the same country. Accordingly, trade unions have taken various initiatives aimed at combating low pay. Overall, these initiatives can be split into 'monetary' (related directly to pay and income) and 'non-monetary' categories.
The trade unions' main focus in terms of pay and income is on the level of the minimum wage and the improvement of the economic situation of low-wage workers through collective bargaining.
In many countries – such as Austria, Belgium, Denmark, France, Germany, Greece, Ireland, Norway, Spain and the UK- trade unions are seeking increases in minimum wage levels. In countries with a national minimum wage system, there are widespread calls for increases in the general minimum rate - as in France, Greece, Ireland and Spain - as well as specific demands. For example, unions in some countries concentrate on young workers, with Dutch unions seeking a higher minimum wage for those under the age of 23 and a number of UK unions with many members in low-paid sectors (eg the GMB general union and UNISON public services union) campaigning to extend the national minimum wage to young workers. In Belgium, the focus for the Confederation of Christian Trade Unions (Confédération des Syndicats Chrétiens/Algemeen Christelijk Vakverbond, CSC/ACV) is maintaining automatic indexation of the minimum wage and following developments in Belgium’s three most important trading partners (France, the Netherlands and Germany), while the Belgian General Federation of Labour (Algemeen Belgisch Vakverbond/Fédération Générale du Travail de Belgique, ABVV/FGTB) has, since a recent increase in social benefits, argued forcibly for a review of the minimum wage in order to narrow the gap with benefits.
In some countries without a national minimum wage, the establishment of such a system is the aim of some trade unions: for example, in Germany, while many unions are sceptical about a national minimum wage, some of those organising in low pay sectors - the Building, Agricultural and Environmental Union (IG Bauen-Agrar-Umwelt, IG BAU), the Food, Beverages and Catering Workers' Union (Gewerkschaft Nahrung-Genuss-Gaststätten, NGG) (DE9911221N) and the Unified Service Sector Union (Vereinte Dienstleistungsgewerkschaft, Ver.di) - are seeking a national minimum wage of EUR 1,500 a month.
In countries where collective bargaining plays a major role in minimum pay-setting, trade unions in many cases seek to increase the agreed minimum rates (this also applies in countries such as France, Greece and Spain where there is also a national minimum wage). For example, since the early 1990s onwards the Austrian Trade Union Federation (Österreichischer Gewerkschaftsbund, ÖGB) has been seeking a minimum sectorally agreed wage rate, currently EUR 872 a month, and has succeeded in pushing up minimum rates to some extent. In Denmark and Norway, trade unions have specifically targeted low-wage jobs in collective bargaining. In Norway, the largest union confederation, the Norwegian Confederation of Trade Unions (Landsorganisasjonen i Norge, LO), usually argues that low-wage workers should be given priority in central-level negotiations, though specific increases or flat-rate pay rises (such flat-rate increases are also seen as a key method of reducing low pay by Ireland's largest union, the Services, Industrial, Professional and Technical Union, SIPTU). Other unions, however, argue that the focus on low-wage workers in Norway's central-level negotiations has contributed to unfair imbalances in wage setting. Finland's current national incomes policy agreement (FI0012170F) contains a special pay increase to be used at sectoral level to improve the position of low-paid workers.
In some of the countries considered, trade unions also focus on fiscal and parafiscal issues. In Belgium, for example, the unions (and other social partners) are targeting a reduction in the social security contributions paid by low-paid workers (BE9910308N). In Ireland, the current national tripartite agreement, the Programme for Prosperity and Fairness (PPF) (IE0003149F), contains a promise to take all workers earning the minimum wage out of the tax net entirely, so that they pay no income tax, while the SIPTU union has focused on tax measures in order to improve the living standards of lower to middle income earners.
In a number of countries – notably France, Greece, Ireland, Italy, the Netherlands and Spain - trade unions have also identified a number of other factors not directly related to pay rates and income levels as important in the fight against low pay
In France and Greece, trade unions seek more restrictive regulations on the use and nature of part-time work as a way of addressing low pay. Similarly, in Italy trade unions are pressing for more rights and protection for workers in 'non-standard' forms of employment, which they consider as usually involving low pay, among other disadvantages, while in Spain trade unions call for a substantial reduction in the country's high level of temporary employment. In Ireland, the unions see improvements in public service provision - such as housing, health, childcare, and public transport - as vital for improving the position of the low paid. Irish unions have also highlighted the issue of the implementation of the national minimum wage, arguing that there are not enough minimum wage inspectors to ensure that employers are implementing the statutory minimum rate.
Trade unions and other employee representative bodies in Austria and Belgium have called for re-examination of the current definitions of low-wage workers, which they see as inadequate in assessing the phenomenon. For example, in Austria the Salzburg Chamber of Labour (Arbeiterkammer) has recently proposed an alternative definition of low-wage work/working poor, using the concept of 'precarious welfare'- which means that a given income does not necessarily provide sufficient opportunities not only to obtain a certain list of goods, but also to participate actively in social life, including the real possibility for individuals to improve their social situation.
The view of employers' organisations on existing minimum wage systems vary to some extent. In countries with a national minimum wage, this seems to be accepted to a large extent, or at least not actively opposed, in many cases - examples are Luxembourg, the Netherlands and Spain (where the low level of the minimum wage may be a contributory factor). The recently introduced national minimum wages in Ireland and the UK, though opposed by employers' organisations prior to their introduction, do not now seem to be matters of major controversy for them. However, the Movement of French Enterprises (Mouvement des entreprises de France, MEDEF), France's main employers' organisation, has been actively campaigning for the abolition of the SMIC minimum wage, calling instead for the government to provide a top-up to the income of employees whose wages are too low (FR0208102F).
Despite this degree of acceptance, where national minimum wages exist, employers' organisations often oppose any increase in their rates - as, for example, in Belgium, Luxembourg and the UK. In countries with no national minimum wage, the introduction of one is likely to be opposed by employers - an example is the Confederation of German Employers' Associations (Bundesvereinigung Deutscher Arbeitgeberverbände, BDA). A key concern of employers is, of course, to keep labour costs down in order to foster business competitiveness
In countries where minimum wage rates are set solely or largely by collective bargaining, employers' organisations are in some cases in favour of a relaxation of the current system, arguing that this would help in terms of employment and business competitiveness. German employers are in favour of a greater use of 'opening clauses' in collective agreements, allowing the payment of wages below collectively agreed rates, and of relaxed pay conditions for employing long-term unemployed people. They also oppose the extension of collective agreements setting minimum pay rates. In Norway, although employers' organisations have been involved in the country's solidaristic incomes policy, they tend to argue that a less rigid wage-setting system would ease labour market entry for groups that face problems in finding jobs.
As well as less rigidity in collective agreements, employers' organisations in some countries propose a number of measures which they state would improve the situation of low-paid workers and increase employment levels. These include notably: cuts in social security contributions (as in Belgium, Germany and the Netherlands); tax and benefit measures to improve the income of the low paid (as in Germany, Ireland, Italy and the UK); relaxation of labour market rigidities and promotion of labour market flexibility (as in Germany and Ireland); and measures to improve employability (as in the UK).
Labour market policies
The evidence presented above indicates the importance of minimum wage arrangements that reduce the incidence of low pay while, at the same time, not having adverse effects on employment. However, the persistence of low pay and the relatively high proportion of low-paid workers across the countries examined implies that the minimum wage can only form one element in any policy to combat low pay and poverty. Government employment and labour market policy also plays an important role in this area. In almost all countries examined, governments have implemented a range of policies which have an impact on low-wage employment.
This impact may be indirect, through the application of general employment promotion policies. However, some policies may be aimed directly at low-paid work. As noted above, low pay is closely associated with low skills, low productivity and little work experience. Some policy-makers and - international organisations (such as the OECD) believe that these characteristics reduce the employment prospects of low-paid workers, acting as a barrier either to better-paid jobs or to entering the workforce in the first place. In this respect, a number of recent changes in government employment and labour market policies have sought to combine a number of instruments in order to improve the employment prospects of low-wage workers. However, it can be argued that such measures often have side effects, notably in terms of a further segmentation of the labour market into groups facing different working realities in terms of employment conditions or pay.
There is a broad and rather homogeneous range of relevant employment policy instruments in the countries examined. Apart from Luxembourg, Norway and Sweden, all the other countries have introduced various initiatives ranging from employment creation and labour flexibility policies to employment subsidy policies, as well as fiscal and parafiscal measures.
In terms of general employment policy, it is widely reported that the jobs created under employment creation schemes are over-represented among low-wage work. In addition, the policies aimed at increasing flexibility, mainly in the form of part-time and temporary employment, all tend to favour low pay, though the real impact on the incidence of low pay is difficult to calculate and varies considerably across the countries. For example, in the UK employment policies have largely focused on supply-side measures, for example 'welfare to work' initiatives. The aim is to reduce the numbers on benefits, by improving the 'employability' of the long-term unemployed, through advice, training and job search facilities. The 'New Deal' for young people (UK0002155F) has elements of compulsion, in terms of offering a choice between education, training and employment. These measures are aimed at those at the lower end of the labour market and, thereby, have an impact on increasing the supply of low-wage workers.
With specific reference to low pay, a common initiative in almost all countries is the introduction of targeted reductions in social security contributions, while in a few countries wage subsidies and special tax credits and benefits apply. However, only from a few countries (such as Belgium and Finland) is consensus reported on the existence of a significant increase in employment as a direct result of reductions in employers’ social security contributions or special tax measures. In other countries there is no overall agreement on the importance of these measures.
In Germany, for example, in early 2002 , the government announced that the 'Mainzer model'- an experimental scheme which provides limited subsidies towards the low wages of former long-term unemployed people and those with a low level of qualifications - would be extended to cover the whole of Germany, having been previously tried out on a pilot basis (DE0005260F). However, academic research is still ambiguous about the long-term effects of such models on the labour market, with one study concluding that there was no empirical evidence so far to demonstrate that nationwide introduction would make a noticeable contribution to increasing employment. Indeed, a Federal Ministry of Labour and Social Affairs on the effects of the pilot projects concluded that after one year they had fallen short of expectations.
Although there is no common definition of low wages across the EU and Norway, and research in this area is not always of great value in comparative terms, it seems that the incidence of low pay varies between 10% and 20% of employees in most countries. Defining low-wage employees as those whose monthly wage is less than 60% of the median national salary, Eurostat puts the EU average at 15%, or nearly one employee in seven. In some countries the incidence of low pay is retreating (such as Norway and Portugal), although it often remains at relatively high levels, while in others it is progressing (eg Austria, Belgium, France, Germany, Ireland, Italy and Spain). Although it is difficult to make comparisons between low wages across EU and Norway, it seems that the situation in the Nordic countries, Belgium and Portugal is somewhat better than in other countries.
National statistical data on the working poor in Europe are scarce but Eurostat statistics put the average proportion of 'poor' employees across the EU at 8%, with national rates varying between 4% and 10%. While the evidence suggests that low-wage employees are considerably more likely than the total population of employees to be poor, the great majority of low-paid workers are not 'working poor'. It thus seems that there is not a causal link between being a low-wage worker and a member of the working poor. An important factor in the relative independence of these two situations in all the countries examined is that the low wages received by one member of a household are frequently balanced by wages received by other members of the same household. Social institutions also play a role: in northern Europe, state institutions and the redistribution of income through state intervention (eg social transfers, notably housing benefits or childcare-related allowances) reduce poverty among low-wage workers; while in southern Europe it is the family that tends to protect low-paid workers from poverty.
The key overall factors contributing to low-paid work are part-time work and low hourly wage rates. However, behind these factors are issues related to the individual characteristics of low-paid workers, the sectors and establishments in which they work and they types of job they do. In sectoral terms, there is a higher incidence of low pay among workers in the private sector - notably service sectors such as the wholesale and retail trade, restaurants and hotels, transport and storage - than in the public sector, although low-paid jobs are also found in health, education and public administration, while in some countries much local government employment is low paid. Occupations with a high incidences of low pay include jobs in sales, personal services, catering, care, cleaning, security and many other unskilled occupations. Small establishment size is very strongly linked to low pay. As well as part-time work, fixed-term, temporary and seasonal employment are also associated with low pay. In terms of personal characteristics, low pay is strongly linked to gender, age, martial status, education/skills and length of service. Women are more likely to be low paid, as are young people, single people, those with a low level of education and employees who have not been in their job for long. Trade union presence and strength, along with the existence of collective bargaining and its level, play a part, as do other factors affecting the balance of power between employers and employees. However, not all of these factors contribute to low pay in all countries, each country having its own particularities.
It is widely reported that low-wage jobs are most commonly found in low-productivity activities or occupations. However, the causal link (low productivity resulting in low wages) often cannot be easily verified since productivity levels are not rigorously defined and measured, especially in the services sector.
An important theme associated with low wages is earnings mobility. The OECD, for example, argues that low-paid jobs often lead to better-paid jobs and thus may serve as a useful entry point to the labour market for less skilled and less experienced workers. However in only a few countries is there evidence that low-paid work often leads to better paid jobs - more commonly, there is evidence that a large proportion of low-wage workers remains in the lower part of the wage distribution, at least in the medium term.
Institutional arrangements overall influence the extent of low pay. In particular, minimum wages set by either law or by collective agreement - which exist in all 16 countries examined - seem to reduce the incidence of low wages, although the benefits of a strong trade union bargaining position or minimum wage legislation are not equally distributed between men and women or between full-time and part-time workers. The effectiveness of minimum wage as a means to improve the pay situation of low-wage workers depends on factors such as the level at which the minimum wage is set and the extent of its coverage. Furthermore, the national evidence indicates neither a positive nor a negative correlation between the level of the minimum wage and the employment of low-paid workers.
Some policy-makers and international organisations believe that the characteristics of low-paid workers reduce their employment prospects either as a barrier to better paid jobs or as a barrier to entering the workforce in the first place. In this respect, and given the persistence and the relatively high share of low pay, employment and labour market policies in many countries seek to improve the employment prospects of the low paid. Apart from minimum wage arrangements that are strongly supported by trade unions, such changes in employment and labour market policies involve a combined approach including wage subsidies, reductions in social security contributions, tax benefits and labour flexibility policies. However, there is no agreement so far on the importance of these measures, while their impact on the incidence of low pay is difficult to estimate and varies considerably across countries. (Elias Ioakimoglou and Eva Soumeli, INE/GSEE-ADEDY, and Mark Carley, SPIRE Associates)
Annex. The coverage and quality of low-wage statistics
- Austria. Official wage statistics are generated by several public or semi-public institutions: the Association of Social Security Providers (Hauptverband der Österreichischen Sozialversicherungsträger, HSV); the Labour Market Service (Arbeitsmarktservice, AMS); Statistics Austria (Statistik Austria); and the Federal Ministry of Finance (Bundesministerium für Finanzen, BMF). Some university departments and private research institutions also collect survey data on income. The Statistics Austria wage statistics do not take account of differences weekly/monthly working time. The report on earnings published biannually by Statistics Austria on behalf of the Court of the Comptroller (Rechnungshof) is helpful when it comes to breaking down the income data for both employees and employers by gender, age, profession and occupation. For the purpose of further and more detailed wage comparisons between subgroups of the population, the data are far from sufficient.
- Belgium. There are several statistical sources dealing with incomes and income levels. First, the Ministry of Employment and Labour calculates the index of collectively-agreed pay on the basis of collective agreements concluded in sectoral joint committees. Second, the National Office for Social Security (Office National de la Sécurité sociale/Rijksdienst voor Sociale Zekerheid, ONSS/RSZ) collects pay data as it has responsibility for collective social security contributions - these data are currently available by application only. Lastly, the National Statistics Institute (Institut National de Statistiques/Nationaal Instituut voor de Statistiek, INS/NIS) conducts surveys on salary structure - the most recent available publication dates back to 1995, but the Institute has been organising annual surveys since 1999.
- Denmark. All enterprises with more than 10 employees are legally obliged to report information about the wages, actual working time, type of work and labour costs of individual wage earners to Statistics Denmark (Danmarks Statistik), which publishes data on wages, working time, labour costs etc broken down by gender, age, education and socio-economic status. The collection of data take place in cooperation with the employers' confederations. The wage data collected by Statistics Denmark is detailed enough to compare individual wage levels. The Danish National Institute of Social Research (Socialforskningsinsituttet, SFI) was responsible for carrying out the household panel survey in Denmark for the European Community Household Panel from 1993 to the end in 2001. Statistics Denmark is involved in the construction of a new set of Eurostat statistics on income and living conditions, which will include data on low wages.
- Finland. Wage comparisons are made in the structure of earnings statistics, collected and published by Statistics Finland. Statistics are collected on a yearly basis and cover all public sector employees and those private sector employees who work in companies employing at least five persons.
- France. The main data source that enables comparisons to be made between levels of individual wages is based on the annual declarations of social data (Dιclarations annuelles de donnιes sociales, DADS), completed by employers. These statistics, covering private and semi-public sector employees, are regularly used by the INSEE statistical institute and are the subject of an annual publication. Wage inequalities are measured only among full-time workers. There are also statistics on state employees. A survey on wage structure is carried out periodically by INSEE, covering full- and part-time workers. Data on low-wage workers can also be obtained from the INSEE labour force survey, which covers all types of employee.
- Germany. The annual wages of individual workers are assessed by the '1% sample survey' conducted by the Institute for Employment Research (Institut für Arbeitsmarkt- und Berufsforschung, IAB) of the Federal Employment Service (Bundesanstalt für Arbeit, BfA). The data are currently available from 1975 to 1995. This survey is the biggest and most reliable source of low-wage statistics. Other relevant surveys are the 'income and consumption sample survey' (Einkommens und Verbrauchsstichprobe, EVS) conducted by the Federal Statistical Office (Statistisches Bundesamt, destatis) and the Office's quarterly earnings survey (in industry, the wholesale and retail trade, and credit institutions and insurance businesses) and structure of earnings statistics (the most recent data are for 1995). The 'German socio-economic panel' (Sozialökonomisches Panel, SOEP), carries out an annual survey, organised by the German Institute for Economic Research (Deutsches Institut für Wirtschaftsforschung, DIW), which includes individual and household income data.
- Greece. The main source of data allowing comparisons between the levels of wages of individual workers is the structure of earnings survey conducted by the National Statistical Service of Greece (ESYE). The survey was last conducted in 1995 and the next is planned for 2003. It covers all sectors of manufacturing and most of the services sector, and enterprises in the private and the semi-public sector. The ESYE also regularly publishes annual figures on changes in wages by sector of production.
- Ireland. There is quite a large body of statistics available that makes it possible to compare wage levels of individual workers. The main body responsible for collecting the statistics is the Central Statistics Office (CSO). The CSO provides the most reliable source of statistics. The independent Economic and Social Research Institute (ESRI) also regularly collects reliable statistics on wage comparison issues. The data available is of a high quality.
- Italy. The two main sources of individual data on incomes are: the survey on income and wealth of Italian households (SHIW), conducted by the Bank of Italy; and the National Institute for Social Insurance (Istituto nazionale di previdenza sociale, Inps) social security archives. The Bank of Italy survey covers all sectors and job positions and examines factors such as gender, age, marital status, level of education, residence, age at first job, main job and secondary jobs. The Inps archives cover all private sectors workers that contribute to compulsory social security funds. Information, especially on wages and employment contracts, is very detailed. Analyses are generally carried out on an annual panel which is made of some 100,000 individual records.
- The Netherlands. The Central Statistical Office (Centraal Bureau voor de Statistiek, CBS) is the main source of wage data. Statistics are collected on a regular basis and are publicly available. Furthermore, the Labour Inspectorate collects data on the observance of the obligation to pay workers the legal minimum wage. The quality of data is excellent and correlations are possible.
- Norway. Statistics Norway (Statistik Sentralbyrå) provides wage statistics that monitor the average income level and average pay increases for groups of employees, by sector, gender and occupation. The wage distribution (dispersion) within these groups is not included in these statistics, but figures on the percentage share of low-wage workers referred can be obtained from Statistics Norway's living conditions surveys.
- Portugal. Wage data are available mainly from the Department of Labour, Employment and Vocational Training Statistics of the Ministry of Social Security and Labour (Departamento de Estatística do Trabalho, Emprego e Formação Profissional, DETEFP–MSST). These data are reliable and representative, covering 71.8% of all employees and 73.4% of all firms in 1998 (though they do not include civil servants, temporary rural workers and the agricultural sector).
- Spain. The main source of information on pay in Spain is the survey of wage costs, previously called the survey of pay in industry and services, conducted and published quarterly by the National Statistics Institute (Instituto Nacional de Estadística, INE). In 1995, INE also conducted a survey of pay structure. Other sources include the National Accounts, which can be used to calculate the average annual wage by sectors.
- Sweden. The National Mediation Institute (Medlingsinstitutet) is responsible for overall wage and salary statistics, and publishes the latest figures every second month. The figures are produced by Sweden Statistics (Statistiska Centralbyrån, SCB) and are based on data provided by employers’ associations. The statistics cover all sectors and are broken down by sectors and/or professions, gender, education and age groups. The trade union confederations, individual trade unions and employers’ associations keep their own wage statistics for bargaining and other internal purposes. The SCB also conducts regular labour force surveys.
- UK. There are two main sources of wage statistics - the New Earnings Survey (NES) and the Labour Force Survey (LFS). Since 1992 the data have been made available quarterly. The Office for National Statistics is responsible for collecting this data. The wage data from the LFS can be correlated with the individual characteristics of the worker and to a more limited extent the nature of the job. The NES is more limited and can be used only in relation to occupation, age, industry, gender, hours and region.