You can now buy the complete works of G.K.Chesterton for ''1.97. A few years ago, they would probably have cost hundreds of pounds and lots of effort. I say this as a counterweight to the Telegraph's claim (via) that the cost of everyday goods has "rocketed" since 1982.

True, many such goods have. So much so, in fact, that in some cases they've outstripped wages. Back in 1982, the average pre-tax weekly wage would have bought 148 pints of lager. Today, it buys 146.

And yet the works of Chesterton aren't wholly atypical either. Back in 1982 no money would have bought you a Playstation or iPod. Their prices have fallen from infinity to affordable - that's an infinite rate of deflation. And things we paid for back then, such as some newspapers and music, are now free. Again, massive deflation.

All this should seem trivial. But it has some under-rated implications.

First, such huge changes in relative prices make it impossible to calculate long-run inflation rates accurately. There's no such thing as a "true" inflation rate. How do you compare the prices of goods that exist today with the prices of goods that didn't exist 30 years ago?

Paradoxically - given that he's considered the father of macroeconomics - Maynard Keynes was more aware of this than this followers.The concept of a general price level, he wrote, is "vague and non-quantitative" and "very unsatisfactory for the purposes of a causal analysis": "two incommensurable collections of miscellaneous objects cannot in themselves provide the material for a quantitative analysis."

Secondly, insofar as people think about the general price level, the availability heuristic can lead them to over-estimate its rate of increase. We buy beer, food and petrol every week and so their price rises loom large in our mind. But we buy books and gadgets only occasionally and so their deflation rate is less salient. This could generate a bias to exaggerate overall inflation.

Thirdly, the general price level is that faced by an average consumer. But many of us aren't average. If you spend less than average on deflation-prone goods, and more on those whose prices rise over time, the cost of living for you might rise more than CPI inflation rates would suggest. For this reason, it is possible - only possible, as there are other factors at play (pdf) - that uprating welfare benefits in line with the CPI represents, in effect, a cut in benefit in real terms.

Fourthly, the fact that the price of technology has collapsed relative to (say) beer over the last 30 years is not just a technical fact about relative prices, but a cause of social change; people respond to incentives, remember. Some of this change might be good - if, say, video games have reduced crime. Other aspects of it, however, might be more ambiguous. I fear (this is just a hypothesis) the decline of pub-going and rise of social networking might be contributing to a hollowing out of the social sphere, in which weak ties - those between neighbours - weaken whilst stronger ties (those between like-minded folk) strengthen. In this sense, worrying about the precise inflation rate is missing the point of what price changes do.