Leave the Analysis for the Experts

My colleague, James Dunning, suggested that I should look into (again) the way that the Swine Flu issue is being reported. In examining the way that the reporting is being done, I'm very grateful to Ben Goldacre who has a long running campaign against poor reporting of statistics in the press.

This week, the UK press has been obsessed with the figure of 65,000 deaths from swine flu. An article in the "Guardian" gives a wonderful example of how the figures have been used by journalists who either do not understand statistics or who don't care in the face of a good story. I really do recommend the story as a case study in the use of statistics.

There is a wider point to be made, I believe. The "Guardian" story shows the importance of using experts - or at least using people who understand the data. Sadly many businesses - and in my experience, most law firms - have the same "good enough" attitude to their own financial reporting, and certainly to their budgeting. A straw poll of law firms (well the four who answered my questions on the understanding that they would not be identified) showed that all of them were going to budget for a substantial reduction in revenue for the 2009/2010 budget year. Well that sounds quite sensible, so far. The range of reductions was from 10% to 30%.

Let me just emphasise that - one firm was budgeting for a 30% drop in revenue. One of my posts earlier last week shows an average of 2.8% increase in revenue for a sample of firms between 2007/2008 and 2008/2009. None of the firms I polled were able to provide any data or sensible explanation for the reduction they proposed.

This is the law firm equivalent of the tabloid press shouting "65,000 deaths!". The various finance committees or boards (or whoever decides these things) seem to have noticed there was a downturn and decided that their business would almost disappear. One firm was willing to speak anonymously in detail. This firm was looking at a budget of 25% less revenue - in the face of proposals from the business units which showed a maximum of 10% reduction.

Data is very useful - firms should have as much useful data as possible. It is vital, however, that the analysis is done by someone who understands what data means; who understand caveats about statistical significance; and who can explain what the data actually means to non-experts. This is vital to both law firms and the tabloid press.