When 20% equals 5 times. How can you lie to people with data and get away with it
--
I always think about data as something that can only tell you the truth. Of course, there are times when we are finding stories and correlations in data that are not necessarily there. But there are a lot of dirty tricks, and one of them is changing the baseline of your axis in a chart (from zero to infinity). Simple trick almost as old as humanity itself that is overused by our politicians and local televisions trying to share their profound data insights about the economy, life, and everything else.
I mean, why would you do that? It took me approximately 30 seconds to find a politician who is using charts with modified baseline. You can take a look at just a couple of examples below.
“We’ve increased the minimum wage almost 4 times”, oh wait:
21% increase in our economy of export that looks like a 5x increase:
When 20% decrease looks like an 80% decrease (length of court proceedings in days):
Or Fox News presenting what happens if Bush tax cuts expire — how to show 4.6 percentage points difference as something five times higher:
Please don’t change the baseline for your axis unless absolutely necessary, so you are not cheating your data users as politicians love to do. Of course, it depends on the context, and sometimes it makes more sense to change the y-axis baseline to tell a better story. Just be aware you shouldn’t be bluntly lying with your data as you can see in the examples above. It’s not a problem of y-axes, it’s a problem of people lying (I recommend short video from Vox below).
If you want to deceive your users, go for it, but I would much rather tell honest and compelling stories with data. I’ll write about those soon.
