Focus on TREF’s Meeting of 28 May 2021


5YR: A Most Important TREF Topic

The reports produced in the context of each 5-yearly review of employment conditions (5YR) on comparisons between CERN’s financial and social conditions and those of comparator employers are among the most important documents the Tripartite Employment-conditions Forum (TREF) examines and discusses. Indeed, these reports form the basis for Management proposals regarding modifications, if any, to CERN’s financial and social conditions to ensure that the Organization is able to attract from all Member States staff of the highest competence and integrity, and thereafter is able to retain and motivate them. These reports, submitted by Management, are thus always subjected to the scrutiny of Member State and Staff Association TREF representatives. In the context of the 2021 5YR (5YR2021), this was the case at the TREF meeting of 28 May 2021.

The first report examined pertained to data relevant for staff member grades 1 to 3. As the 5YR method prescribes, the relevant comparator data should be collected from “the employers established in the local region of the Organization (Geneva, Vaud and neighbouring France) that offer salaries that are among the most competitive”. The report prepared by CERN’s HR Department on the basis of consultant data indicates that the employers which offer salaries that are among the most competitive are established in Switzerland (Geneva and Vaud) and that CERN salaries are in line with those. This report did not call for any specific comment from the Staff Association.

Procedural Deviations of Importance

The second report covered data relevant to the remaining staff member grades (4 to 10). It was commissioned by CERN’s HR Department from the International Service for Remuneration and Pension (ISRP) of the OECD. Its first conclusion was fully in line with those arrived at in previous 5YRs i.e. that “the ISRP identified Switzerland as the geographical market offering the most competitive salaries amongst CERN’s Member States [and] Germany […] as the second most competitive market”. The data on which this conclusion is based were, as far as can be ascertained, obtained and analysed in a manner fully compatible with that relied upon in previous 5YRs.

However, much to the surprise of the Staff Association, data relevant for other conclusions presented in the report were obtained and analysed in a significantly different manner from those of the previous 5YRs. This resulted in conclusions which were also quite different from those of previous 5YRs. In brief, the key differences are:

  1. CERN’s salaries being defined net of taxes[1], one has to apply taxes to the comparators’ salaries before they can be compared to CERN’s. While Zurich’s taxation rates[2] were chosen for this purpose in respect of Swiss comparators during the two previous 5YRs (2010 and 2015), Geneva’s were chosen for 5YR2021. In its 5YR2015 report, the ISRP indicated that “in order to keep the same approach used in the past study, and because it is the largest city and the most important economic center, Zurich has been chosen as the reference for taxation purposes”. Now, this year, ISRP indicates that “given that CERN headquarters is located in Geneva, this location has been retained as the reference canton and commune for taxation purposes”. But, the ISRP does not explain why its reasoning suddenly changes so significantly (Zurich does remain the “largest city and the most important economic center” in Switzerland) nor why, suddenly, “keep[ing] the same approach used in the past study” is no longer important! The impact of the Zurich-Geneva change is very significant: about 6 percentage points on average i.e. a large fraction of the difference between CERN and comparator salaries as measured in the 2015 5YR!
  2. More than one third of the CERN salaries used as a basis in the 5YR2021 comparisons are significantly higher than those used in 5YR2015: by 14-15%, no less! But, we know our salaries have not increased so much in the intervening years, be it for those of us in the corresponding functions or not! What is at play here is the fact that, for these functions, the salary range of the corresponding MERIT benchmark jobs (BMJ) was increased compared to the corresponding MARS range (the third grade’s maximum salary is higher than the corresponding maximum in MARS). Yet, this only changes the possibility for some of the colleagues in these BMJs to attain such higher-than-before salaries in the latter part of their career. This is all the more a mere possibility as promotion to the third grades is not at all as ‘normal’ as the promotion from the first to the second grades. In fact, if a fourth grade had been added above the third and access to it had been made even less ‘normal’, 5YR2021 would have shown CERN salaries even higher than in 5YR2015, but again without any of us (or potential recruits) having seen their (resp. offered) salary increase. Thus, 5YR2021 CERN salaries do not reflect the reality, which debases completely the comparisons.

To summarise: as a result of the first effect, comparator salaries are artificially lowered and, as a result of the second effect, CERN salaries are artificially increased. The two effects combine to produce completely artificial comparisons!

No Real, Factual Explanations

The ISRP attempts to explain away the much smaller differences between CERN and comparator salaries by referring to changes in purchasing power parities (PPP): “The increased competitiveness of CERN salaries in comparison to past studies is mainly due to [the fact that] the cost of living in Germany has increased steadily as part of PPP calculations, while on the other hand, the cost of living in Switzerland has comparatively decreased at a faster rate”. But, firstly, as Swiss comparators pay salaries in Swiss francs as CERN does, PPPs only impact CERN-Germany comparisons. Changes in PPPs cannot explain the so-called “increased competitiveness of CERN salaries” with respect to Swiss comparators! Secondly, this change in (Germany-Switzerland) PPPs is only ~3% whereas between 5YR2010 and 5YR2015 it was ~10% and did not have such an impact on the comparisons! So, the ISRP’s explanations necessarily fail to explain the observed marked change in the comparisons.

In addition to the above, it should be mentioned here that repeated Staff Association requests for explanations by ISRP of specific differences and changes in the comparison procedure and data between 5YR2021 and 5YR2015 were set aside because no time-series data were initially requested or because the window for comments/questions to the consultant hired by ISRP had closed. Both might be true, but one expects 5YR methods to be stable and transparent, and the Staff Association representatives needed extra time to put a finger on extra-hidden effects. Given this, it would only have been normal for Management to agree to further investigations and for ISRP to undertake them. Much regrettably, not so.

Unprecedented and Momentous Conclusion

All in all, for the above reasons, the Staff Association had to take the unprecedented step to not accept the results of the comparisons as determined by ISRP and endorsed by Management.

The consequences of this and the hiccups in the 5YR2021 data and procedures described above cannot be determined as we write this article, but they might be momentous. So, stay tuned!

Note: Information on the comparisons relevant for Fellows and Associated members of the personnel (MPA) will be presented in a future ECHO article.


[1] Hence the internal taxation mechanism.

[2] As noted by the ISRP, “taxation rates in Switzerland differ amongst existing cantons and communes”; hence the necessity of picking a place (commune) of taxation.