Who knows best? Thousands of ordinary people with the time and patience to answer a long-list of questions from an opinion pollster? Or the painstaking, professional statisticians who measure inflation by a variety of ingenious formulae every month?
The Bank of England, which has in its time expressed dire doubts about other official numbers, has little doubt about these. The gentlemen in Whitehall really do know best.
Yet the question arises because the Bank itself commissions a regular quarterly survey by NOP starting with the critical, factual question what "best describes how prices have changed over the last 12 months". This regularly shows a substantial balance of opinion convinced that prices rise materially faster than the official indices.
This is a huge survey involving nearly 4,000 respondents - four times those taking part in many dodgy election-time polls, NOP picks them from 350 randomly selected spots across the country, then weights the results to match the demographic profile of Great Britain. The phrasing of the questions is scrupulously neutral. This is not the way you get a freak result.
The same goes, surely, for the elaborate efforts of the boys and girls at National Statistics to get it right.
Yet in February, for instance, when NS said the Bank had hit its two per cent target spot on with the Consumer Price Index, after under-shooting for several months, and the old Retail Prices Index was running at 2.4 per cent, 39 per cent of NOP's sample put inflation at three per cent or more. Sixteen per cent plumped for five per cent-plus.
Only 20 per cent of them got the right bracket, two to 2.99 per cent. Of the rest, 27 per cent under-estimated it and 14 per cent confessed they had no idea.
That leaves a decisive proportion of real people spending their own real money, paying real bills, convinced that prices rise faster than the statisticians say and the Bank believes.
There are clever answers. You can argue that for years inflation has been held down by falling prices for TVs, audio systems, computers - arguably for cars, too, in that airconditioning adds less to the price now than a sun roof did a few years back.
These are all "big-ticket" items that relatively few people buy in any one year. So there is always a built-in majority that has not recently benefited from the lower prices.
A less innocent explanation is that, however scrupulous the number-crunchers, Governments place the goalposts for them.
All over the world Govern-ments like low official inflation. It saves them money - as employers, and on state pensions and benefits.
In America, they use a measure of "core" inflation that leaves out the cost of food, for heaven's sake, as well as energy, on the grounds that they are volatile.
A couple of years back our own Gordon Brown brazenly switched his target to an index that leaves out all costs of owning your home, including council tax - but home-buyers count them in when the opinion pollster calls.