My son, who watches his pennies, was sure the eggheads in Canberra wouldn’t have noticed “shrinkflation” – reducing the contents of packets without changing the price. No. This trick’s intended to fool the unwary punter; it doesn’t fool the statisticians. It counts as a price rise.

But now for the two reasons the CPI can indeed be misleading. The first is that averages can conceal as much as they reveal. Remember the joke about the statistician who, with his head in the oven and his feet in the fridge, said he was feeling quite comfortable on average.

The most recent news that, according to the CPI, prices rose by 0.8 per cent in the three months to the end of June, and 6 per cent over the year to June, was an average of all the households – young, middle-aged and old; smokers and non-smokers, drinkers and teetotallers, no kids and lots, renters, home buyers and outright owners – living in the eight capital cities.

Now note this. Economists, politicians and the media tend to treat the CPI and the “cost of living” as synonymous. But if you read the fine print, the bureau says that, while the CPI is a reasonably accurate measure of the prices of the goods and services in its metaphorical basket, it’s not, repeat not, a measure of anyone’s cost of living.

Why not? Partly because it does too much averaging of households in very different circumstances, but mainly because of the strange – and, frankly, misleading – way it measures the housing costs of people with mortgages.

The cost of being a home buyer is the interest component of your monthly payments on your mortgage.

Loading

But that’s not the way the CPI measures the cost of home buying. Rather, it’s measured as the price of a newly built house or unit. Which makes little sense. Many people with mortgages haven’t bought a new home.

And even those people who did buy a newly built home, did so some years ago when house prices were lower than they are now.

The bureau changed to this strange arrangement a couple of decades ago. Why? Because the Reserve Bank pressured it to. Why? Well, as you well know, the RBA uses its manipulation of interest rates to try to keep the annual rate at which prices are rising, as measured by the CPI, between 2 and 3 per cent on average.

But, after it had adopted that target in the mid-1990s, it decided that it didn’t want the “instrument” it was using to influence prices – interest rates – to be included in the measure of prices it was targeting, the CPI.

So, the bureau – unlike other national statistical agencies – switched to measuring home buyers’ housing costs in that strange way. And the bureau began publishing, in addition to the CPI, various “living cost indexes” for “selected household types”.

The main difference between these indexes and the CPI is that home buyers’ housing cost is measured as the interest they’re paying on their loans, not the cost of a newly built house. But, of course, different types of households will have differing collections of goods and services in the basket of things they typically buy.

So, whereas the CPI tells us that prices rose by 6 per cent over the year to the end of June, the living cost indexes show rises varying between 6.3 per cent and 9.6 per cent.

Among the four selected household types (which between them cover about 90 per cent of all households), the type with the highest price rises was the employees, whose costs rose by 9.6 per cent overall.

That’s mainly because most of the people with mortgages would be is this category. Mortgage interest charges rose by 9.8 per cent in the quarter and (hang onto your hat) by 91.6 per cent over the year.

Loading

At the other end of the spectrum, supposedly “self-funded retirees” had the lowest living-cost increase of 6.3 per cent – mainly because almost all of them would own their homes outright.

Then come age pensioners, with cost rises of 6.7 per cent – few with mortgages, but some poor sods renting privately.

And finally, “other government transfer recipients” - those of working age, including people on unemployment benefits, on the disability pension and some students. They’re costs are up 7.3 per cent. Some of these would have mortgages, most would have seen big rent rises.

What this proves is that using interest rates to control prices makes the cost of living worse before making it better.

Ross Gittins is the economics editor.

Ross Gittins unpacks the economy in an exclusive subscriber-only newsletter every Tuesday evening. Sign up to receive it here.