The jitter plot that could make us slightly less jittery about inflation

Isn’t that picture pretty? Half the reason for posting at all is the data-art enabled by the detailed inflation figures. The other half is the useful information that the more granular view can convey. But let’s first explain the chart, before we get into the “so what”.

The chart is a bit of a hybrid between a time series — it shows the months from January 2019 to January 2022 from left to right — and a so called jitter plot (or jitter chart). A jitter plot is where you “jitter” the datapoints left and right within a particular column so that you can see more of them; it’s particularly useful when you have a lot of data with the same coordinates which would otherwise be overlapping in the chart and hard to see. The jitter basically allow us to visualise more of the datapoints.

I’ve also colour coded the CPI items (more on them here) based on the estimated labour intensity — as this should give us some indication as to whether we might be headed for a wage-price spiral and hence more probelmatic, and endemic, inflation in the economy. On this graph, a dark red bubble indicates a high-labour-intensity item, such as social care or restaurants and bars. A dark green, in contrast, indicates a low-labour-intensity item, such as fossil fuels or chemicals. The size of each bubble is its weight in the CPI index.

So, other than being pretty, what is this chart telling us? Well, if you focus in on the last few months, then there are three different things happening:

1. We’ve seen big price increases (dots high up on the graph) in some very low-labour-intensity (green) items, such as petrol, diesel, natural gas, and electricity. The prices of these items are set (broadly speaking) in global markets, so UK wages will have very little impact on future inflation here.
2. We’ve seen most of the dots move above the zero line — so most items increased in price between January 2021 and January 2022. This is in stark contrast to, say, during 2020, when the same dark green dots tended to be below the line, i.e., exhibit big drops in prices.
3. And we’ve seen the more stable, labour-intensive red and orange dots in the middle of the graph edge gradually up over time, but at a fairly moderate pace. In January 2022 (last column), most labour intensive services (such as child care, education, and restaurants and bars) saw price increases in the range of 3–5%, relative to January 2021.

I hope you’ve enjoyed learning what a jitter plot is and how it can be helpful in visualising patterns in granular data. And I’d more than welcome your thoughts on whether this analysis of inflation is useful!