What’s up with inflation? The question really is, what is not down?

The big economic story at the moment is inflation, whether it will continue apace, what central banks will do about it, and so on. I wanted to get a better handle on what’s actually going on with the underlying prices — not least to get a sense of how broad-based or not the recent price increases have been.

The Consumer Price Index (the measure of inflation I’m focusing on) is actually made up of price increases and weights for around 220 different items, ranging from rental costs (weight of around 9%) to book binding services and e-book downloads (weight of around 0.01%). As you would expect, there normally is a wide distribution of price increases and decreases across these items. The overall average therefore hides a lot of detail.

The chart illustrates this by looking at December inflation (year-on-year price increases) in three different years: one low-inflation (2019), one moderat-inflation (2017), and one high-inflation (2021). Even in the high-inflation year of 2021, when the annual inflation recorded in December was 5.4%, there were several items — such as mobile phones, motor vehicle insurance, and pets — whose prices declined relative to December 2020.

Looking at the distributions (where the height of the bar measures the number of items in each bin and the width of it is the sum of the weights of those items in the CPI), it seems that the high inflation of 2021 is as much due to small number of product categories where prices have declined, as it is about product categories where prices have increased.

Indeed, even in December 2021, on the basis of simple category numbers, there were more items that had their prices increase by a modest 1–2% than there were items in the 4–5% bucket. Granted, the weight of the items in that latter bin was higher, contributing to an overall higher average for the month. The latter included, for example, the weighty (5%) category of “restaurants and cafes”.

Over the years (the detailed data is available from 2016 onwards), it is pretty remarkable how many products have seen consistent decreases in prices. In the 72 months in my dataset (among the 220 categories), there were more than 40 items where prices went down year-on-year in more than 50% of the months in the dataset. For example, prices of personal computers declined in 64 of the 72 months analysed. For poultry, declines were evident in 59 months, for frozen veg (excluding potatoes) in 56 months, and so on.

At the other extreme, there were also products and services which have seen consistent price rises. In more than 20 categories, prices went up every month in the time period analysed. Some of these categories included fast food and takeaway services, education, and passenger transport. Not surprisingly, consistent price rises have been much more prevalent in domestic services than in internationally traded goods.