This paper presents three main limitations. First, really understanding the differences between BERD and the Scoreboard would require matching the two data series at company level, but the BERD micro data are not publicly available. Second, the paper stays at a very general level and there is little expert knowledge about the relevance of the items discussed in previous sections. For example, looking at sectors of performance, it would be quite possible to discuss which are the most relevant boundary problems concerning the private sector. Third, some of the differences in section 2 could be reasonably supported by evidence: for example, differences between the calendar and financial year could be better justified if explored country by country.