Methology and FAQ
What does the indicator ‘days not worked due to industrial action, per 1,000 employees’ mean?
The days not worked due to industrial action, or the strike volume, can be generally defined as the indicator that is the result of multiplying the number of workers involved – either directly, indirectly or both – by the duration of the industrial action for any strike or lockout.
This result is divided by the number of people in employment and then multiplied by 1,000 in order to correct for the number of wage-earners and to make comparisons over time and between countries. Although, from a power relations perspective, lockouts should in principle be separated from strikes, most data do not make a distinction between these two types of industrial action. The number of days not worked due to industrial action per 1,000 employees is considered the most reliable indicator for making historical and country comparisons, although systems for collecting data on industrial action can change over time and are not identical between countries.
How reliable is the data on days not worked due to industrial action in general?
The data on the days not worked due to industrial action underestimates the actual number of days not worked. There are several reasons for this underestimation, including changes made in the collection methods, the systematic under-reporting of strikes and the deliberate exclusion of strikes from the statistics. Particularly ‘small’ strikes, i.e. of a short duration with only a few workers involved, simply go unnoticed. Although such small strikes do not substantially influence the number of days not worked, they can still be important. Furthermore, some statistical bureaus have set a formal threshold for collecting the data on industrial action or deliberately do not collect data on certain types of strikes, such as general strikes or strikes in certain regions or in certain sectors such as the public sector.
Finally, it should be noted that the data only reflects the number of observed cases of industrial action and that the data at the aggregated level masks sectoral differences and does not report on any other collective or individualised expressions of conflict at work.
See also the article by Dribbusch and Vandaele, here.
What does the indicator ‘workers involved in industrial action per 1,000 employees’ mean?
The workers involved in industrial action, or the strike size, can be defined as those workers actually taking industrial action, by being on strike or taking part in a lockout, and who are thus ‘directly’ involved in the action. Workers may also be ‘indirectly’ involved by being laid off as a result of the industrial action. The strike data often includes these ‘indirectly’ involved workers, making estimations about the actual propensity to strike less accurate. The number of workers involved in industrial action is divided by the number of people in employment and then multiplied by 1,000 in order to correct for the number of wage-earners and to make comparisons over time and between countries.
What does the weighted average mean?
The weighted average is needed to correct for the number of people in employment. So, countries with a large number of people in employment contribute more to the final average of days not worked than countries with a small number of people in employment. The weighted average is more accurate than the simple average for assessing the total number of days not worked due to industrial action in Europe.
Why do some countries not have data on days not worked due to industrial action?
We do not really know why a number of countries have recently stopped collecting data on industrial action. This decision has been formally made for ‘austerity reasons’ in several countries. However, the real underlying reasons probably include political authorities’ ‘sheer embarrassment’ about a high strike volume, their increasing disinterest in industrial action, their fear of being less attractive for foreign direct investment, or purely ideological motives. Furthermore, Eurostat, the statistical bureau of the European Union, no longer collects data on industrial action.
Why is there a mix of different sources used for certain countries?
A mix of different sources has been used in cases where there were only partial data on industrial action available or if no official data were available at all.
How is the number of days not worked due to industrial action calculated for France?
In France, two sources are available on the days not worked due to industrial action. DARES provides data on days not worked in the private sector, based on a survey of companies with 10 employees or more, although only the days not worked per 1,000 people in employment is made publicly available. The data from the Direction générale de l’administration et de la fonction publique provides data on days not worked in the public sector and some semi-public enterprises. There is thus strike data for ‘fonction publique de l’Etat’ but no data for the ‘fonction publique territoriale’ and the ‘fonction publique hospitalière’ (except for the year 2008 in the latter case). The share of employment in the public sector is calculated by dividing the combined total employment of the three public sectors (fonction publique de l’Etat,fonction publique territoriale and fonction publique hospitalière) by the total number of people in employment – both national sources are used for the numerator and denominator. This share of about 22 per cent is used to determine the rate of employment in the private sector, which is then used to determine the days not worked in the private sector. The days not worked in the public and private sector are then added together and divided by the total number of people in employment (source: Eurostat). All in all, the days not worked due to industrial action are probably still underestimated, as the number of days not worked in the fonction publique territoriale and fonction publique hospitalière is not known (except for 2008).
Do you have recent data on industrial action in any of the following countries: Bulgaria, Croatia, Czechia, Greece, Italy, Luxembourg or Slovenia?
As far as we know, there is no recent data available for Greece, Italy and Luxembourg. With regards to Bulgaria, Croatia, Czechia or Slovenia no official data is available at all.
What are the other sources used for this strike map?
Information on the regulation concerning strike action has been retrieved from the ICTWSS database and the ETUI working paper on ‘Strike rules in the EU27 and beyond’. The classification of the collective bargaining systems is based on P. Marginson and C. Welz (2014), Changes to wage-setting mechanisms in the context of the crisis and the EU’s new economic governance regime, Dublin, Eurofound. The peaks in the number of days not worked due to industrial action are interpreted with the help of articles from the European Observatory of Working Life.
Is there other or alternative data on industrial action or social protest?
Yes, there is other and alternative data on strikes and social protest. Based on the articles of the European Observatory of Working Life – EurWORK of Eurofound, and the European Protest and Coercion Database, Kerstin Hamann (University of Central Florida), Alison Johnston (Oregon State University) and John Kelly (Birkbeck, University of London) created a dataset covering general strikes in Europe. In addition, Stefan Schmalz and colleagues of the Friedrich-Schiller-Universität Jena created the Jena Conflict Data (JenaConDa), a database of press coverage of social conflicts in Europe.