Is inflation avoidable?

Inflation in the UK is at a 40 year high, with prices rising across the economy. Shoppers can look for cheaper products, but many are already buying the lowest-cost options and the prices of those bargains are going up faster than those of higher-priced goods.

Inflation has defined the summer in the UK, with prices having risen by more than 10% over the past year and pressure on household budgets predicted to intensify (see Figure 1).

Some claim that inflation is unavoidable and that it is inherently unequal since poorer households often have fewer options in terms of shopping and essential items take up a bigger share of their incomes. Others suggest that by looking around for bargains, even the cash-strapped should be able to avoid the worst price rises.

This article uses the latest data to test these two points of view: is inflation avoidable?

Figure 1: The UK's consumer price index, 1988-2022

Source: Office for National Statistics (ONS)

The UK’s three measures of inflation – the consumer price index (CPI), the consumer price index including housing (CPIH) and the retail price index (RPI) – are closely related. They are all currently over 10% in the UK.

These measures are the country’s most essential economic yardstick – since they are hard-wired into policy decisions. Inflation is used to uprate pensions, public sector pay and to set the pay-out on inflation-protected debt.

Inflation also influences the Bank of England’s decisions in interest rates, and the payments made and received by savers and borrowers across the economy. The official figures are based on records for over 100,000 prices, which are collected each month by the ONS. Knitting together these underlying historical files (known as ‘price quotes’) results in a database of over 40 million prices going back to 1988. (More detail on the data is available in this paper and the files can be found on my site, here.)

The argument that high prices can be avoided by judicious shopping rests on the common-sense observation that for any given item there will be many options – and many prices – on offer.

To take a late summer example, consider a trusted gardener’s tool: the garden spade. So far this year, the ONS has already collected almost 1,000 prices for garden spades across the country. These range from £5.99 to £44.99: £14.99 is the most common (modal) price; £19.99 is the middle of the distribution (the median); and the average (mean) price is £21.73 (see Figure 2).

Of course, there is likely to be a difference in quality between a humble spade bought for less than £10 and a fancy one for over £40. But the basic economic function of the product is the same.

If all we cared about in the economy was spades, then the ‘thrift’ argument would be right. Massive savings are possible if you go for bargain basement goods.

Figure 2: Price of different garden spades

Source: ONS

But we care about more than garden tools, and so to measure the cost of living in a meaningful way, we need to track a bundle of items. To simplify comparisons over time, I focus here on the prices of goods that were included back in the 1988 CPI and are still used today. There are 106 of these perennial items and a table summarising them is available here.

This bundle includes a range of products – food and drink, tobacco, homewares, clothing, healthcare products and services – that people across the UK have been buying for the past 34 years.

To follow how shoppers of various types – from those that opt for the cheapest options, to those that pay for luxury – we can track the distribution of prices over time. Figure 3 identify nine spending levels, corresponding to evenly spaced points on the price distribution.

In 2022, the group opting for the lowest priced bargains (buying at the 10^th percentile) could get the entire bundle for £1,650. Those buying the highest-priced options (the 90^th percentile) would need to spend over £7,000. Those in the middle group (the 50^th percentile) would spend £3,500.

The fact that the bundles include one-off purchases – pushchairs, for example – that have a wide range of prices magnifies the spread. There is a wide range of prices for many goods, and so there are savings from ‘trading down’ to comparable, but lower priced, items.

Over the long run, the lower end of prices has risen less quickly, as might be expected, with the long-run inflation rate for our best bargain bundle close to 2%. Again, this seems to support the intuitive thrift argument: over the long run, the price of the cheapest goods tends to rise slowly. (This price distributions for all of the goods can be seen in this visualisation).

Figure 3: Consumer bundles by price decile, 1988-2022

Source: ONS

High costs for low prices

So, should the spread of prices in the economy make us less worried about inflation? There are three reasons why not.

The first is that for many goods, inflation is impossible to avoid. We can see this by comparing the plots below for garden spades and driving lessons (Figures 4A/4B).

Over the past 30 years, the spread of spade prices has changed very little: you could buy one for £15 in 1990 and you can today. But for driving lessons, the pattern is different. As time passes, prices rise at the top and disappear at the bottom. So, while driving lessons priced at £11 per hour were common in 1990, the cheapest today is closer to £20.

Learning to drive can be an economic and social necessity – if you live rurally, for example – and the only option is to pay higher prices. Energy markets, covered extensively on the Economics Observatory site, are another example of this.

Figure 4: Price evolution, 1990-2022

Panel A: Garden spades

Panel B: Driving lessons (one hour)

Source: ONS and author's calculations

The second problem is that you can’t keep switching to ever cheaper options. At some point those driven to find bargains will end up with much of their shopping the bargain bucket. And while historically, prices here have risen more slowly, that pattern has evaporated with the UK’s latest bout of inflation.

In fact, those that buy goods in the cheapest three buckets of goods (the 10^th, 20^th and 30^th percentiles) are seeing higher rates of inflation than those at the top (the 70^th, 80^th and 90^th percentiles). Families that shop at the bottom now face higher inflation than those that shop at the top.

Figure 5: Average inflation rates, consumer price baskets

Source: ONS and author's calculations

Note: Buckets defined as average across inflation rates for decile groups (Bottom = 10^th, 20^th, 30^th; Middle = 40^th, 50^th, 60^th, Top = 70^th, 80^th, 90^th).

The third problem is volatility. To see why this matters, step back and consider why inflation matters. In part, it is due to the costs that workers and shops face – often referred to as ‘shoe-leather’ and ‘menu’ costs – when trying to keep up with rising prices.

But it is unexpected shifts in inflation that are especially damaging, since these undermine plans in the economy – from wages rates to investment decisions. This is why John Maynard Keynes saw unanticipated inflation as the most malign type, writing about the ‘precarious life of the worker’ and ‘excessive windfalls’ to profiteers (Keynes, 1956).

As the UK’s inflation targeting regime was set up, volatility and the violation of expectations were seen as a vital target. Useful, and very readable, articles from this period include Leigh-Pemberton (1992) and Briault (1995). Price rises are supposed to be low, and predictable.

Assessed over short windows of time, prices are highly volatile, as a series of seminal studies focused on US data by, Emi Nakamura and Jón Steinsson (2008), Pete Klenow (2010), and Virgiliu Midrigan (2011) have shown. Some of the reasons why are intuitive – firms offer temporary sales and when multiproduct companies decide to change prices, they often change them all.

The important point here is that prices, as well as trending up with inflation, are in constant flux. And the problem is that the price bundles for the perennial goods shown here seem to have a lot of volatility at the bottom. The long-run standard deviation (the distance from the average) of the cheapest food prices is around 11% higher than for the most expensive, for example.

Other categories, including services, show volatility at the top and bottom, but less in the middle. The concern is that those shopping at the low end are likely to pay a high price in terms of inflation uncertainty.

In summary, there are elements of truth in both perspectives on prices. A capitalist economy provides choice, and we have seen a huge ‘fanning out’ of prices in the UK over the past 30 years. So, many families will be able to make savings by trading down – opting for bargain basement options and forgoing luxury. Indeed, this intuitive response to higher prices seems to be playing out across the economy, as evidence from supermarkets and consultancies shows.

But this does not reduce the problem, nor ease the headache for the Bank of England and HM Treasury. You can’t scrimp forever, and the analysis here shows that, once you reach the bargain basement, prices are rising fast and are volatile. The latest inflation in the UK, the worst since the 1970s, is one in which those at the bottom are paying some of this highest costs.

Technical notes/data access

The data used in this essay were first published as Davies, 2021. I update the databases each month and make them openly available via my website here. Please get in touch if you have questions related to a research or policy project.

The specific numbers used in these figures, together with the JSON spec that draws the charts (using Vega), are all available on my GitHub page, here.