In my previous organisations I have used the HSE Stats as a benchmark, purely because they were the enforcers in both cases and the stats were "Official Government Figures". However, I see no problem with using both HSE and LFS figures as benchmarks (a seperate line on the graph for each) providing your management don't start thinking they're 'allowed' more accidents than they are currently having. Do be aware of the following headline though, which hints at under-reporting in some sectors: http://www.shponline.co....il-riddor-irregularities

BOTH the HSE and ONS Labour Force Surveys are state-of-the-art products. ALL professionally-designed (and other) surveys contain margins of error: the professional ones are those where the authors can tell you reliably the extent of the error. Error is part of the very nature of this kind of measurement - and it's worth ensuring that management understand why or they are not measuring their production or quality validly either. If you want better-focused external benchmarks, see what the trade associaions and employers' organisations in your sector may have on offer. Some of them are well designed and state of the art.

Yossarian You've actually illustrated the very point I made, namely about 'SOME' trade associations The example you chose is one in which MANY trade associations and employers organisations are not staffed with relevant specialists. If you wish to test my assertion, access relevant data from the CBI and interpret them validly. Some professional sociieties e.g. CIPD, are also reliable sources of valid data on behaviour at work. The major issue concerns the qualtiy of statistical evaluation necessary to make valid judgements about out behaviour at work; for example, to interpret the changing pattern of absence from work due to injury and other causes, it's both the CBI and CIPD data on absence are needed alongisde those of the HSE and the ONS over a period of years.