Why statisticians don't think your iPhone is more expensive than you

One of our favorite things to write for FT Alphaville is that reader reviews often lead us to look at the topics we write from a completely different perspective.

With that in mind, we want to focus on responding to last week's post about the central bank's perception that the gap between inflation and public behavior is widening. Some readers have argued that statisticians look at the cost of technological progress differently from you or me when measuring prices.

Compared to last year's model, Joe Volkswagen may pay more for this year's iPhone, with pure inflation, while statisticians will reflect part of the increase as a manifestation of innovation (for example, better cameras), And discounted for this.

Although the price is higher, this usually leads to higher prices for new but better quality products, which are recorded as deflation.

We also consider it worthy of further investigation, so for most of Friday, we have carefully studied the UK National Bureau of Statistics' technical manual to understand its inflationary methods to get an accurate picture of what is going on.

We emphasize continuous quality because we believe this is the key to understanding what is happening here.

Consider the improvement in television quality over the past decade. Ten years ago, the average setting in the average home might have been smaller, the sound was worse, and SmartTV features were much less (if any). Obviously, these innovations have an impact on the quality of experience. But if the price goes up by a few hundred pounds, how much should it be discounted?

ONS looks at this problem through "direct quality adjustment", which is described as an "attempt to place quantitative values ​​on quality or specification differences". Their statisticians use two methods to assess the value of direct quality adjustments: hedonic regression and option costs.

The first picture shows that a new computer costs £ 550, while an old computer costs £ 475. However, once technological improvements are taken into account, the benchmark price (ie the representative cost of the old machine used to measure inflation) will become £ 562.40. This means that introducing new models is deflationary. Accurate to 2.2%. In contrast, without regression, the inflation rate attributed to the new PC was 15.8%.

Here we can see the logic made by statisticians. However, we believe that monetary policymakers can control the public's perception of inflation in some way by setting inflation targets, which is another challenge to this myth.

Few would see a new computer cost £ 75 more than the deflationary price of an old computer. But this is exactly what the Bank of England is targeting (and used to anchor public expectations).

We don't think Apple will stop producing new models of the iPhone in the near future, even if the innovations introduced are becoming more and more fake and cyber traps in nature. So how can the UK and other central banks hope to close the gap in inflation expectations?