The Office for National Statistics will shortly be publishing new figures on inflation rates for different types of household. They will also be updating the method for collecting individual prices from supermarkets. However, anyone hoping for evidence that inflation is much higher for low-income households is likely to be disappointed.
In particular, the anti-poverty campaigner Jack Monroe has tweeted that the ONS ‘have just announced that they are going to be changing the way they collect and report on the cost of food prices and inflation to take into consideration a wider range of income levels and household circumstances’. This isn’t quite right.
There are three things going on.
First, on Friday the ONS will resume publication of data which estimate the inflation rates for different types of household. These are broken down by income, expenditure, housing tenure, retirement status, and whether or not there are children in the household.
This is nothing new. The last figures were released in February 2020, then the publication was suspended due to Covid. The past reports have also not shown large differences in inflation rates between different types of household, including income groups. In most cases the gaps have only been a few tenths of a percentage point.
There is a widespread assumption that the pandemic has had a far bigger impact on the inflation rates for poorer households, mainly because they spend a larger proportion of their income on food and, especially, domestic energy bills. Friday’s data will shed more light on this.
However, the rise in inflation has been far broader. There have also been large increases in the prices of cars, motor fuel, holidays, recreation, household goods, and hospitality. These are all areas where richer households typically spend a larger proportion.
The upshot is that overall inflation rates probably still only differ by only a few tenths of a percentage point across different income groups.
The second change is that the ONS will be updating the way that data on shop prices are collected. Currently, these are mainly drawn from surveys by market researchers. In future, the ONS plans to make more use of a much larger sample of electronic data based on real transactions at checkouts.
This method should be more accurate. However, this isn’t really news either. It is part of a long-term transformation programme, and all the changes ‘just announced’ have been set out before, including in a briefing published in November last year.
It is also important to understand that the impact on the official measure of inflation could go either way. My own guess is that the current method results in inflation being over-estimated.
This is because the prices actually paid will often be lower than the data collected by an ONS representative walking round the shop. Real customers will switch to cheaper alternatives if one item is more expensive, take advantage of special offers, and shop around.
There are some factors that might pull in the other direction. When prices are high, poorer people might only be able to afford smaller packs, where the unit cost is higher. Properly measured, this is an additional type of inflation, sometimes known as ‘shrinkflation’.
It has been also suggested that we need a ‘Vimes Boots Index’, named after the character in the Terry Pratchett ‘Discworld’ novels who could only afford cheap boots that needed replacing more quickly. This is similar to the concept of ‘shrinkflation’, only applied to quality as well as quantity. Alternatively, this index could focus on the costs of ‘essentials’, i.e. non-discretionary spending.
Gathering more information here is clearly a useful exercise, but the ONS already has this in hand. Indeed, it already publishes data broken down into discretionary and non-discretionary items. In November 2021, the 12-month inflation rate for non-discretionary spending was 4.1%, which was actually lower than the 5.2% for discretionary items.
The third change planned by the ONS is to improve their information on which goods and services are bought by each household types. This is also a continuous process – the basket of goods and services is frequently reweighted – and not the result of lobbying by any particular activist. The better availability of granular data, due to new technology, is the real driver here.
That said, Monroe and others are right to draw attention to the additional pressures faced by poorer households. The headline CPI is inevitably an average, concealing a wide range of different circumstances.
Even if inflation rates are not much different now, the main upside risk in the coming months is likely to come from energy prices – and especially what happens to the Ofgem price cap on domestic bills in April. This could indeed hit those on lower incomes particularly hard.
Poorer households are also more likely to spend a higher proportion of their incomes on essentials, giving them less flexibility to manage their budgets. And they are unlikely to have large savings that they can dip into to maintain spending during a temporary squeeze.
This means that, even facing the same rate of inflation, the living standards of low-income households may be at much greater risk when prices surge. But there is still little evidence that inflation itself is much higher for poorer people.
This is an extended version of a piece first published by The Spectator on 27th January 2022
ps. the data published by the ONS on 28th January confirmed that those assuming inflation is higher for low-income households are barking up the wrong tree.
Plain-speaking Economics 