Did you hear the one about the Chief Executive who asked a senior analyst, “How much of our total salary spend goes direct to women and ethnic minorities?”
The request was marked urgent and when the analyst was rebuffed by HR on grounds of data protection, he had a bright idea. Googling the government statistics he uncovered a mine of information and was quickly able to find a breakdown of average salaries earned by the various social categories within the relevant sector. After a few calculations and some nifty powerpoint a report emerged containing the numbers that the CEO had asked for. No questions were asked about the data sources or assumptions and the company set about building these statistics into its CSR strategy and action plans. It was only three years later (by which time the CEO had become chairman, and the analyst had moved on to a more lucrative position at a competitor) that questions arose about why the positive measures the company had adopted were having such limited impact on the bottom line. An unpaid intern discovered the cause and was duly given a grilling by the new CEO.
This fictional but plausible vignette illustrates the problem with using industry statistics to estimate emissions using “input-output” methods.
The key problems are:
- input-output tables give an average allocation of national emissions by economic sector per unit turnover or employee. However, your business is unlikely to fit neatly into the median or mean sectoral profile.
- national level data collection processes are crude and coarse compared with your internal systems, and do not correspond correctly to the accounting standards applied to businesses for emissions and other impacts.
- aggregated sector level data sets will not capture improvements you make to the business, your supply chain, or the specific circumstances of your operations.
So, while input-output analysis can give a broad picture of the order of magnitude of potential impacts it is not a suitable approach for monitoring and reporting of impacts.