5.5.3 Levels of compliance

As stated under section 5.3.1, in the underlying study three different kinds of levels of compliance must be distinguished to add to the answering of the main research question. Table 5.3.1 above shows and explains those different kinds of levels. Table 5.5.3 shows the a-levels of compliance per country per year.

Table 5.5.3 Levels of compliance per country and year1

Table 5.5.3 shows that, overall, the percentages are very high and at first glance reassuring, but there is more to these figures. Regarding level I.a (the number of provisions the company claims to apply) an average increase of 0.93 per cent from 2005 to 2007 can be seen, with Germany (+1.59 per cent) and Italy (+2.28 per cent) showing the highest increase. Level III.a (applied and sufficiently explained) also shows an increase from 2005 to 2007 by 0.3 per cent, with again Germany (+0.93 per cent) and Italy (+1.54 per cent) showing the highest increase. The analysis shows that - as desired by among others the EU (EU Green Paper 2011) - the code compliance increases, as well as the quality ofthe explanations in the case of non-compliance. Nevertheless, it is simultaneously noticeable that the increase in code compliance levels off (for Belgium and the Netherlands a slight decrease can even be seen). Possibly the attention paid to code-compliance and improvements in this matter decreases or the highest point of code compliance has been reached (for further discussion see section 6.4.4). This last proposition seems unlikely since 1.52 per cent of the explanations in 2007 (567 explanations) are still insufficient or no explanation is provided. Because of the choice for the comply or explain principle within Europe and the 'one size does not fit all' approach it can be defended that some deviations (e.g. due to interim situations) will continue to exist, but nevertheless explanations that are understandable and legitimate should be aimed for. As stated in subsection 5.3.1, level II.a ought in any case to increase to 100 per cent, since it concerns ' applied and explained' (compliance and deviations with explanation), as does level III.a (applied and explained sufficiently).

To be able to research the levels of compliance I.b, II.b and III.b (see table 5.3.1) the code provisions comparable between the five countries must be distinguished since a comparison per code provision is not possible between the five countries. Table 5.5.3a shows the 13 main categories of code provisions to be distinguished (see for the subcategories Annex III). The categories in Italics (1, 4, 5, 8, 11 and 13) are not present in all corporate governance codes and therefore left out of the comparison and determination of the levels of compliance Ib, II.b and III.b. For category 7, Italy cannot be taken into account as regards the period of 2005 since code provisions on the remuneration of supervisory board members were not present then. Most of the topics not present in the code or present in only some countries are explainable by culture, legal origin or the focus of the code as explained further in chapter 4. For instance, Belgium has a two-tier but no one-tier board structure and therefore pays no attention in the code to the one-tier board structure (see section 4.3 - in practice, a hybrid board structure). Moreover, sometimes matters are provided for in annexes to the code, such as the internal auditor in Annex C of the Belgian code.

Table 5.5.3a Code subjects for b-levels of compliance

Table 5.5.3b below shows the levels of compliance based on the comparable code provisions per country and year (level I.b, II.b and III.b) for the code provisions on the management board (general provisions and provisions on remuneration), supervisory board (general provisions and provisions on remuneration and independence), the three key committees and shareholders (categories 2, 3, 6, 7, 9, 10 and 12 in table 5.5.3a).

When compared with table 5.5.3 (the a-levels), the scores for the convergence levels (the b-levels) are, overall, slightly lower. Since the scores are in percentages the ratios are the same. The difference is, however, explainable by the fact that apparently the provisions which are part of the convergence comparison (categories 2, 3, 6, 7, 9, 10 and 12 in table 5.5.3a) are the provisions not complied with most of all, as a result of which the convergence scores are slightly lower. Again, it is simultaneously noticeable that the increase in code compliance levels off and for Belgium and the Netherlands a slight decrease can even be seen. The increase in compliance during the years 2005 until 2007 is slightly higher for the convergence comparison (0.94 per cent compared to 0.93 per cent for level I) with again Germany and Italy showing the highest increase. As hypothesised in subsection 5.3.2, an overall tendency towards more compliance and convergence between the countries under review can be seen. Taking the levels of compliance III.a and III.b into account, an improvement in the quality of the reasons provided for non-compliance is noticeable as well.