Flexicurity – Remember the flexicurity model that was launched by the European Commission in the mid 2000’s? Claiming that there existed such a thing as a ‘golden triangle of flexicurity’ (see illustration below), the Commission urged member states and trade unions to give up on job protection in exchange for adequate unemployment benefits and active labour market policies.
The inspiration for this was clearly found in Denmark with the country being hailed as the perfect illustration of how a flexible labour market with low restrictions on employers to fire workers could still offer high security of employment.
Source: OECD Employment Outlook 2004
In this context, the latest OECD Employment Outlook from the summer of 2013 is extremely interesting. In this publication, the OECD thoroughly reviewed its database on job protection indicators. In particular, job protection arrangements resulting from collective bargaining practice and case law have now been included more systematically than was the case before. This implies that these new OECD indicators should provide an improved picture of reality, in particular when weak job protection in labour law is corrected by collective bargaining agreements and/or case law imposing additional and more stringent job protection.
Flexicurity, what flexicurity?
This review of the OECD database leads to surprising conclusions, in particular regarding the system of flexicurity in Denmark. The graph below shows the newly estimated employment protection indicators for regular (open ended) contracts, thereby also adding job protection in case of collective dismissals. The value of the employment protection legislation (EPL) indicator in the graph is obtained by applying weights to these two sub indicators, with regular contract protection counting for 70%, and collective dismissal protection for 30%. Indicator values close to zero indicate a very low level of job protection, whereas scores going up to an indicator value of 5 or 6 point to employers experiencing extreme difficulties in firing workers.
Protection of regular workers against individual and collective dismissals, 2013
The graph shows that Denmark does not have a labour market that is particularly flexible. With an indicator value of 2.3, the level of job protection in Denmark is not below but exactly at the OECD average. It can also be seen that regular contracts and collective dismissals in Denmark are protected at a level that is similar to Germany, Spain and Greece. Also, Danish job protection is not very much below the levels of job protection registered in France and Italy. Meanwhile, the gap in job protection between Denmark and the flexible Anglosaxon labour markets is significant, with the UK and the US at values as low as 1.5 and 1. In practical terms, the 2.3 score for EPL in Denmark translates into 4 (2) months of advance notification for a white collar (blue collar) worker having 4 years of tenure in the job, with additional notification periods, procedures and delays in case it concerns a collective dismissal.
Delving deeper into the history of the statistics allows us to further back up the conclusion that the OECD has been widely off the mark on this all these years. In the final graph we see the values for the sub-indicators on job protection of regular contracts for the year 2004 as estimated by the OECD back in 2004, (source: the OECD 2004 Employment Outlook). These are then compared with the EPL values for the same year 2004 which are now to be found in today’s 2013 OECD database. From this comparison, it is clear what has happened. Back in 2004 the OECD (the blue bars in the graph) estimated that regular jobs in Denmark were poorly protected with an indicator value of just 1.5. Based on the estimates done in 2004, Denmark could indeed be characterised as having a flexible labour market, with the degree of job protection as low as in the UK and Ireland and substantially below continental countries such Germany or France.
Now look at the red bars, registering the current OECD estimate from 2013 of the level of regular job protection back in 2004. This turns the picture completely around. For Denmark, the new estimate for 2004 now comes out substantially higher, at a value of 2.1. This actually means that, in contrast to the estimates from the earlier database, Danish workers in 2004 were benefiting from a level of job protection that is twice as high as in the UK and is not far removed from the job protection levels of France and Germany.
A similar comparison (not shown here) can be done on the other sub-indicator, the protection of jobs in case of collective dismissals. Its conclusion is that, in case of collective dismissals, Danish workers were as strongly protected as German workers and even more protected than their French or Italian colleagues.
All of this actually means that the whole policy of flexicurity, as it has been promoted all these years by the European Commission, has been based on a statistical illusion. The argument according to which the success of labour market performance in Denmark can be put down to the fact that workers and not their jobs are being protected is simply not correct. Through its system of collective bargaining, Danish workers are being offered robust levels of job protection. The true peculiarity and advantage of the Danish system lies in the fact that Denmark invests heavily in both passive and active labour market policies. It does not lie with employers having the possibility of easy firing.