Do collective agreements ensure full coverage and compliance with EU Directives?

Denmark continues to witness a debate on whether EU Directives should, or can, be implemented via collective agreements, or through legislation. New research findings on the "spill-over" effects of collective agreements into areas not directly covered by them throw some interesting light on the question. The survey examines overtime working and suggests that collective agreements ensure compensation and limit long-term use, but do not prevent the widespread existence of overtime.

The question of whether EU Directives can and should be implemented by collective agreements or by legislation is again the subject of debate in Denmark. The social partners in both the private and public sectors, together with the Ministry of Labour, are fully in support of the "Danish model" (DK9708122F). This is an industrial relations model characterised by the use of collective agreements as a mean of implementing both EU Directives and national policies which fall within the scope of their domain. The philosophy behind the Danish model is simple: if the responsibility to implement a policy is in the hands of the social partners, then the social partners will ensure that implementing collective agreements will be applied in an efficient and practical manner.

Although employers' organisations in the private sector cover approximately 50% of the labour force and trade unions approximately 80%, the social partners firmly believe that the adoption and application of agreements and the norm-setting effect of the negotiations results, in practice, in almost full coverage. This coverage, according to the beliefs of the social partners, is such that it does not provide a sufficient basis for questioning the value of the Danish model, and therefore does not provide any support for the use of legislation to ensure full implementation of what had been agreed. The social partners fear a development in which the gradual introduction of legislation would undermine the Danish model of collective agreements. This is because the incentives for membership of employers' associations and trade unions would thereby be eroded.

New findings from a research project conducted by Steen Scheuer, a labour market researcher at Copenhagen Business School have recently been published. The main objective of the project was to examine and test the validity of the social partners' argument regarding the actual coverage of collective agreements. This project was undertaken by the use of questionnaires, which asked respondent employees about their coverage by collective agreements and the effects. The 1993 EU Directive on working time was implemented in Denmark through collective agreements, and the coverage and application of this Directive was used as an example for the project.

Coverage of collective agreements

The survey on the coverage of collective agreements showed that only about 52% of respondents in the private sector positively confirmed that they were covered by collective agreements. Since all employees in the public sector are covered by a collective agreement, this adds up to a total coverage of 62% of the entire Danish labour market. This percentage is high by international comparisons but nevertheless low compared to the perception of the Danish social partners. The survey has consequently been cited in support of the view that the assumptions which underlie the Danish model should be called into question.

Although the questionnaire method employed in the project may be questioned, estimates based on factual information suggest that the above figures are relatively reliable. However, what has not been taken into account is the fact that the coverage of white collar-workers within the area covered by the Danish Employers' Confederation (Dansk Arbejdsgiverforening, DA) is higher than appears to be the case in the survey results. It seems reasonable therefore to qualify the survey estimate by assuming that the private sector coverage is approximately 60%, and thus the total coverage for the entire labour market is approximately 75%. This estimate has been produced by theUniversity of Copenhagen's Industrial Relations Research Group (Forskningsgruppen Arbejdsmarkeds Organisationernes Sociologi, FAOS).

The survey results also illustrate that the low degree of coverage is particularly widespread among employees in higher-paid jobs and among other groups of white-collar workers, who by and large enjoy the same pay and working conditions as set out (as a minimum) in collective agreements. However, among blue collar-workers, approximately 25% are not directly covered by an collective agreement. On the basis of these figures, Steen Scheuer argues in favour of always using legislation as the means to implement EU Directives.

The results of the survey were published in 1996 in "Fælles aftale eller egen kontrakt i arbejdslivet. Udbredelse af kollektive overenskomster, faglig organisering og skriftlige ansættelsesbeviser blandt privatansatte", Steen Scheuer, Nyt fra Samfundsvidenskaberne, København (1996). The main findings are published in English in "Collective bargaining coverage under trade unionism: a sociological investigation", Steen Scheuer, British Journal Of Industrial Relations, Blackwell, Oxford, Vol. 35, no. 1 (1997).


A new survey based on the research results from 1994 has examined the extent to which collective agreements have a "spill-over effect" into areas not directly covered by them. Using working time and the amount of overtime as examples, the survey arrives at the following conclusions.

Firstly, collective agreements in reality set the norm for the entire labour market when it comes to working time. The 37-hour working week, which is set out in collective agreements, has a spill-over effect and is almost taken to be a rule applying for the entire labour market. As the survey concludes: "The informal institutional spill-over effect of the collective agreements on a normal working week is considerable, and ensures a general norm on working time in a satisfactory manner - in a manner which is better than in countries which use legislation."

Secondly, collective agreements do not prevent overtime. Although collective agreements state that overtime is to be limited, these provisions are not implemented in practice. Collective agreements do allow for overtime, if it is paid for, and overtime does occur to a certain extent. Accordingly to the survey, 73% of the respondent employees in the private sector stated that they worked overtime from time to time and 44% worked more than eight hours' overtime per month.

Thirdly, the coverage of collective agreements does have an effect upon the amount of overtime and the extent to which overtime is compensated by way of time off in lieu and additional pay. Compared with employees directly covered by a collective agreement, employees who are not covered by such agreements more frequently work more than eight hours of overtime per month. This means that the Danish model does not have an universal spill-over effect with regard to overtime. It is merely those who are directly covered by a collective agreement who can be sure of less overtime and, if they do work overtime, of being paid for it. The survey concludes, however, that a certain spill-over effect cannot be ruled out: 39% of those not covered by collective agreements receive an allowance for overtime, compared with 72% of those who are covered. According to the authors of the survey this can be seen as an expression of a "more informal institutional normative connection, where the norm of the collective agreements has a spill-over effect beyond its factual area of validity".

The collective bargaining system does have a measurable effect, but is not universal. The survey also shows that 4% of respondents work more than the maximum 48 hours per week, which is prohibited by the EU Directive.

The survey therefore suggests that trade unions could use legislation as a means of preventing employers taking advantage of overtime. It is worth noting that the group of employees with the most frequent overtime are those higher up the occupational hierarchy.


The survey is valuable and commendable because it has produced results which challenge traditional assumptions about the Danish model. However, the questionnaire method can be questioned - questionnaires can be problematic when respondents reply to particular questions presented to them, and it could be that some respondents genuinely believed that the work they were carrying out was "overtime work", despite contrary provisions in the collective agreement. It is this type of criticism that the social partners have made of the survey. However, one could argue that the social partners themselves should have participated more actively in the survey by offering more precise statistics on the factual coverage of their collective agreements, their implementation and their spill-over effect.

Trade unions and employers continue to refuse the use of legislation as a means of implementing what they have agreed between themselves, or in the application of EU Directives. The social partners stress the positive results that collective agreements bring. In this connection, one could question whether legislation de facto is more efficient then collective agreements. Can legislation ensure full compliance? The survey points out that this might not be the case, claiming that in Greece and Portugal, where working time is subject to legislative regulation, non-compliance is often the case in practice.

Whether a new balance between legislative and agreement-based regulation is needed depends to some extent upon the ability of the social partners to maintain a high degree of organisation and to ensure a dissemination and spill-over effect to the entire labour market. Whereas legislation by definition covers everyone, compliance may in practice not be a reality. Then the question arises as to how full compliance can be ensured. (Kåre FV Petersen, FAOS)

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

Eurofound welcomes feedback and updates on this regulation