A while ago, I saw a neat chart by Olivier Ballou that he created for the American Enterprise Institute. It shows Consumer Price Index data from the US. And since we just learned about and calculated the Consumer Price Index last week, I thought it’s a splendid idea to update Olivier’s chart with current data this week:
We’re looking at an indexed chart. The chart doesn’t show the absolute prices of TVs or childcare, but the relative price change: Childcare is up 100%, which means it’s double as expensive now than it was in 1998. And the price of TVs is down almost 100%: You can now watch Game of Thrones on a device that costs half of what it cost in 1998 (or, you know, keep watching it on your super pricey Apple device). To learn more more about indexed charts and how they’re calculated, head over to this Weekly Chart I published a while ago.
A chart never shows us the whole picture. We should always ask ourselves: “What don’t we see in this chart?” – especially when the chart gives us the hint that it shows a selection of the data. Two examples:
The chart up there can easily make us think that hospital services and college textbooks are the two US goods for which the price increased the most. But we don’t know that. We only know that, of the shown goods, hospital services and textbooks are the ones with the steepest price increase. Maybe the price of all of the non-shown goods increased even more. We simply can’t know just from looking at this chart.
This chart can also make us believe that all prices increase or decrease quite steadily over time, because that’s the case for all the goods we see. But prices are not necessarily just going up or down in a straight line. If we look at prices that are not shown in the chart – like the one of gasoline and goods from depending industries – we see that the prices can fluctuate quite a bit over the last 20 years:
Olivier Ballou probably decided against including gasoline prices because its wiggly line would have drawn all the attention to it – and away from hospital services and college textbooks. In this case, the selection was a simple design decision. Olivier didn’t have bad intentions to deceive us. And still, we let ourselves be deceived: It’s just too easy to believe that what is true about the selection is also true about the whole. (The same mechanism works hard in our heads when we form stereotypes based on anecdotes.)
I’ll be on vacation next week, but if everything goes according to plan, you’ll still get a Weekly Chart. See you then!