Data limitations

Given that the restructuring events database relies on selected media titles, its coverage of restructuring activity in each Member State is indicative and cannot be considered representative.

In view of size thresholds for case inclusion, the monitor reports almost exclusively on restructuring in medium and larger sized firms; this size bias in turn leads to an over-representation of the manufacturing sector where company size tends to be larger. Variability of national-level media coverage of restructuring events from country to country leads also to country biases. This is reflected in higher levels of ERM reporting in some Member States (e.g. UK, Poland) and lower levels in others (e.g. Greece, Bulgaria).

Also, it should be taken into account that once reported in the media, the decisions announced by the companies may be changed for various reasons without necessarily giving rise to a second press article. We therefore underline that restructuring job losses are as originally announced; depending on the individual case, these figures may or may not tally with the actual final job loss or gain.

In spite of these biases and data limitations, the dataset does generate a picture of labour market restructuring, especially in relation to sectoral restructuring activity that is broadly consistent with data coming from more representative sources such as the European Labour Force Survey. It has also tended to anticipate well overall trends in (un)employment in European labour markets while providing unique data on the proportion of overall larger-scale restructuring-related job loss accounted for by different forms of restructuring (offshoring, internal restructuring etc). Other positive advantages of the restructuring events database are its timeliness, its identification of individual cases of restructuring based on publicly available information and its uniqueness as an EU-wide dataset of larger-scale restructuring events.