Department of Labour, Employment and Social Dialogue 2014 Collective Bargaining in 2013, Reports. travail-emploi.gouv.fr/IMG/pdf/ procedures that affect the duration of agreements and their validity beyond their expiry also affect the coverage of negotiations. Such procedures tend to protect workers if employers refuse to negotiate and strengthen the bargaining position of unions in the event that employers want to lower standards. The aim of the reforms of recent years was to encourage trade unions to negotiate concessions in times of recession. Representation criteria (and, if used, public interest clauses) are intended to reflect as much as possible the situation of a wide range of companies, but cannot take into account their full diversity. As a result, few countries allow exemptions from extensions. In the Netherlands, clearly predefined derogation criteria are even a precondition for prorogation. In addition, companies can apply for an ad hoc exemption from the ministry if they can justify an exemption.27 Hijzen et al. (2019[58]) indicate that 191 applications for ad hoc exemptions were submitted by Dutch companies between 2007 and 2015, but only 58 were accepted. Although there are no formal rules on exceptions in Switzerland, companies with annual turnover of less than 1.2 million francs (about 1.2 million euros) were excluded in 2012 (Visser, 2018). Another way to better reflect the heterogeneity of companies and avoid the “single limit” of extensions would be to promote differentiation within agreements, as is the case in the Dutch metallurgical industry, where two agreements are signed and extended in practice, one for companies with 35 or more employees and the other for companies with less than 35 employees. The French reform of 2017 also conditioned the extension of a sectoral collective agreement to a differentiation of content between large and small enterprises.28 Note: The international comparability of strike data is influenced by different definitions and measures.

Many countries exclude small work stoppages from their official registers and use different thresholds for the number of workers involved and/or the number of days lost. In some countries, strike statistics may also exclude work stoppages in certain sectors such as the public sector (such as Austria, Belgium, France, Germany, Latvia, Portugal and Turkey) or a particular type. B strikes, such as Chile, Costa Rica, Estonia, Hungary, Israel, Korea, Latvia, Lithuania, Mexico, Turkey, the United Kingdom and the United States. Conversely, in some countries, workers who are indirectly involved (i.e. those who cannot work because others are on strike in the workplace), such as in Costa Rica, Denmark, Estonia, Finland, France, Hungary, Ireland, Lithuania, the Netherlands, New Zealand, Poland, the Slovak Republic, Switzerland, Turkey, the United Kingdom and the United States may include work stoppages due to the lack of equipment provided by striking companies. In general, forms of trade union action that do not involve full stoppages, such as go-slows, silence and other workplace protests, are not included.