is a journalist, author and creator of the BBC Radio 4 programme More or Less
I hear and see on the BBC today, reported in shocked tones, that one in 150 businesses went bust in 2008. This is unequivocally a Very Bad Thing.
I don't hear that this is close to an historic low. I don't hear that in the last recession the rate was worse than one in 40. I don't hear that for the entire period from 1984 - which is the first year for which the insolvency report supplies data - all the way through the 1990s, right up to 2005, it was apparently worse than it is now. That's right, in only two years of the last 24 has it been better. I don't hear that, even if you take the rate for the very last quarter of 2008 and assume it was true of the whole of 2008, you are looking at a liquidation rate lower than for any year from 1984 (and probably before) to 2002 or 2003.
Yes, the number of companies in trouble short of receivership - a carefully chosen slice of all companies in trouble - has gone up sharply in the last quarter of 2008 compared to the last quarter of 2007. But 2007 looks like an extraordinarily good year by historical standards, and the percentage deterioration quoted - 220% - represents just over 1,600 companies, most probably tiny, out of about 2 million.
This is an absolute change of 0.08 per cent, or eight in every 10,000. Is that a big number?
One caveat. It is possible that something odd happened to the data around 2002-04, when the rate of corporate insolvency fell sharply, possibly as a result of legislative encouragement to the self-employed to incorporate; though the insolvency report does not mention this. Even if we take out the fall in these two years, 2007 still looks like an historic low.
This is not to argue that things are wonderful. They aren't. But not being wonderful is not the same as being apocalyptic. When I sometimes say that our business coverage lacks proportion, it is not because of what it says, which is generally true, but what it conspicuously doesn't. The principle seems to be to shove in the worst-possible sounding statistic that you can find without serious attempt to understand it, and to ignore the rest.