According to popular mythology, the overwhelming majority of people who set out to read Stephen Hawkings' book A Brief History of Time never get beyond page on which the first equation appears.

I had a similar experience last month when I was reading an article on the UK's Office of National Statistics (ONS) web site. It explained, in some detail, how that bellwether of the economy, the Retail Sales Index, is calculated. And it is highly impressive: There are discussions regarding statistical outliers, the timing of bank holidays, the impact of prescription sales, the coverage of smaller businesses and all sorts of other variables that might have a small but important impact on the headline number. But, inevitably, as I read deeper, the sudden occurrence of greek mathematical symbols quickly drove me back to the home page.

There was a point to all this. Just before Christmas, I found myself on a training course with a specialist in retail property. Rather than being buoyed by the seasonal retail frenzy, he said his sector was preparing for a nasty recession. "Was this evidence that ‘bricks-and-mortar' retail was beginning to suffer from the shift to online retail?" I asked. He could not really say.

This caused me to reflect on my own major purchases in 2004. In almost every case – whether it was for a digital camera or a fridge freezer – I had gone into one or more local shops to check out the merchandise – and then bought the goods for at least 20% cheaper on the Internet.

This is not unusual behaviour, of course. This was clearly foreseen by the dot-com evangelists back in the late 1990s, and many retailers at the time genuinely feared much of their High Street would be wiped out. Many panicked and over-invested on online initiatives, only to relax later as the whole thing began to subside. To underline the lesson, companies such as the over-indulgent and absurdly over-capitalised Boo.com collapsed in dramatic fashion.

But the shift online has been much more gradual – just not particularly well recorded. This was my reason for digging into the official Retail Sales Index. I felt it important to know just how fast Internet sales are really growing in the UK, and also if the official index, which is one of the figures used by the Bank of England to set interest rates, accurately captures the Internet sales figures.

The recent numbers made for some very interesting reading. In the UK, retailers are coming to terms with one of the worst Christmas seasons on record. Sales in the UK fell 1% from November to December, the biggest fall in that period since 1981, when the monetarist prime minister Margaret Thatcher was applying a tourniquet to the UK's money supply. Today's problem, of course, is mostly explained by the lowering of consumer spending confidence following the slowdown in the property market and by the impact on disposable income of five quick increases in the interest rates.

But split out online sales and you see that Internet sales rose 9.8% in December, a healthy increase that continues a buoyant year for online retail. In the US, Internet sales are rising even faster: in the run up to the end of the year, US online sales jumped 29%, with visitor numbers at some sites increasing faster still. Wal-Mart, belatedly pushing its online business, recorded a 79% increase in web site visitors.

There is, of course, no doubt that the UK's retail recession is real enough and is explained by macro-economic factors. But it also seems certain that Internet sales are being substantially under-reported. Indeed, the Internet is barely mentioned in the thousands of words explaining the ONS methodology for calculating the index.

Internet sales are only recorded if the retailer is one of the 5,000 large and small companies tracked by the index. But many new and small online retailers, and certainly those outside the UK but selling into the UK, are simply not recorded. I bought most of my goods from small retailers who I had never heard of before – and, I think it quite likely, nor has the ONS.

There is one spectacular omission from the figures: downloadable music. Although sales already run into tens of millions of dollars worldwide, music purchased by UK consumers online is still not accounted for in the index – unless sales go through one of the basket's of leading retailers. Yet the loss of CD sales certainly is included. If the index is to continue to be a true measure of retail health, then that basket is going to have to be much more fine-grained – especially as broadband and mobile connectivity make the Internet retail truly ubiquitous.