In one of my recent sessions, “Accidental Data Lies: How Poor Visual Can Choices Mislead”, I take a deep dive into common chart crimes and the ethics of data visualisation and storytelling. One of my favourite examples is truncated axes.
If the term sounds technical, don’t worry, it’s actually very simple. Truncating an axis means omitting the baseline, usually zero, so the axis starts at a higher value. Most of us have seen this many times without realising it. In fact, many tools automatically adjust axes for us, quietly hiding the baseline in the process.
And to be clear, truncated axes aren’t always wrong. They can be genuinely useful when you want to highlight small variations within a narrow range of values. The problem starts when this design choice changes the story the audience takes away. By visually zooming in, we risk giving users a misleading first impression, and first impressions in data matter. They can build trust or slowly chip away at it.
Now if you are wondering why truncated Y-axis may cause a bad first impression, this article is for you!
Understanding The Issue
Truncated axes exaggerate visual differences by compressing the scale. This means that relatively small variations between data points can appear far larger than they actually are. As a result, the audience may overestimate differences between groups and assume one is dramatically outperforming another.
It’s all about perception, and perception is a powerful little thing!
You might remember the infographic that circulated on social media comparing popstars heights. At first glance, it’s funny: the visual gap between the heights looks absurdly large. Unsurprisingly, it went viral on social media.
But before you think that this is just a funny and harmless image, we need to understand that omitting baseline is also one of the most common ways data is misrepresented, and this is frequently used to make one group look better than another.
Misrepresentation is when the data is shown in a way that’s not accurate, so the underlaying data is not wrong. News channels seem to have a particular crush on this method specially when talking about politics or when reporting a supposedly “alarming” rise in something scary, like violence or interest rates.
See a Truncated Y-Axis In Action
Have a look at the chart below. At glance, the column sizes imply that border apprehensions nearly tripled from 2011 until 2013, but the underlaying values only had a 16.4% increase. This is because the Y-axis is truncated making the jump from 165,244 to 192, 298 look more accentuated.
Now look at the same data plotted with the y-axis starting at zero. The data hasn’t changed, but the sense of urgency has. Without the truncated axis, the increase appears far less dramatic, putting the change back into proportion.
This is why column (bar) charts require extra care. When we use them, we’re implying that the length of each column (bar) is proportional to the underlying value. Once the baseline is removed, that visual contract breaks, and so does trust.
If you do need to omit the baseline, be explicit. Clearly state the axis range. Where possible, consider offering both views: one with the axis starting at zero and one without, depending on the insight you want to support. And don’t forget alternative chart types. A line chart, for example, can show change over time without exaggerating magnitude, though, as always, it depends on the story you’re trying to tell.
When Omitting The Baseline Is The Right Choice
Now, this is where things get interesting. I’ve spent a lot of time explaining how truncated axes can mislead, yet there are situations where not omitting the baseline does more harm than good.
Consider the chart below plotting the average annual global temperature in Fahrenheit. In the first chart, the axis starts at zero, exactly as we’re often taught it should. Technically correct, but visually misleading. Zero degrees Fahrenheit is around –17.8°C, far below any realistic global average temperature.
By forcing the axis to start at zero, the line becomes so flattened that it appears almost static. A viewer could easily walk away thinking that temperatures haven’t really changed, or that global warming isn’t a serious issue. Not because the data says that, but because the visual quietly downplays the trend.
Now compare that with a chart below, where the axis starts closer to the actual temperature range. Suddenly, the upward trend is clear and unmistakable.
This is the nuance. Sometimes omitting the baseline is necessary to highlight meaningful variation. When that’s the case, transparency is essential. Label the axis clearly. Explain the decision in a caption or footnote. Help the audience understand that the omission is intentional and informative, not deceptive.
Designing With Intent, Not Just Accuracy
In a data-driven world, being able to understand and interpret data is more important than ever. As you’ve seen throughout the examples shown here, what might seem like a minor design choice, such as adjusting an axis range, can completely change how we perceive the insight. A small visual tweak can turn a modest change into a dramatic story, or flatten a meaningful trend into something that looks insignificant.
This is why accuracy alone isn’t enough. Even when the data is correct, the way it’s presented shapes the narrative people walk away with. Our responsibility as data designers goes beyond getting the numbers right. We need to be intentional with our visual choices and mindful of the stories they create. Because even when the data is right, the story can still be wrong
Thank you for joining me on this journey. Until next time, let’s keep crafting accessible and ethical insights that make a difference!
