Right-Side Y-Axis Labels: Fan or Foe

At the end of the data visualization class I teach at Georgetown University, I show students some classic, historical examples of data visualization, including from Florence Nightingale, John Snow, and William Playfair. This year, something caught my eye in Playfair’s famous “Exports and Imports” line chart—he placed the y-axis labels on the right side of the chart. Playfair even calls attention to the choice in his own caption: “The Bottom line is divided into Years, the Right hand line into £10,000 each.” So, it doesn’t seem an accidental decision he left unremarked—he flagged it as a feature of the chart.

Historical line chart by William Playfair titled 'Exports and Imports to and from Denmark & Norway from 1700 to 1780.' Two hand-drawn lines track export and import values (in units of £10,000) across the decades, with the area between them shaded pink where imports exceed exports, labeled 'Balance Against,' and shaded tan where exports exceed imports, labeled 'Balance in Favour of England.' From 1700 to about 1755 imports slightly exceed exports, hovering between roughly 60 and 100. Around 1755 the lines cross, and exports rise sharply to nearly 190 by 1780 while imports stay near 90, creating a large favorable trade balance for England in the later decades.

Positioning the y-axis labels on the right side is clearly not the norm. I suspect the reason for the left-placement standard is that the English language is read from left to right, and English readers tend to read in F or Z patterns, starting in the top-left and then either reading down the page or across in those patterns.

But there are roughly one to two billion people who use languages that are read right-to-left, like Arabic and Hebrew, and another 150 million people whose language is written vertically, like Japanese and traditional Chinese. Muna Alebri, Noëlle Rakotondravony, and Lane Harrison collected 128 visualizations from articles published in Arabic news outlets and found that designers to to apply right-to-left text approaches more frequently for categorical data, though it was less consistent than they had perhaps hypothesized.

In the English context, placing the y-axis labels on the right side of a chart—especially a line chart—may be a way to bring readers’ attention to the most current data rather than where the data originated. For example, in the Bloomberg chart below on China’s fiscal spending, do we care more about how much the government spent 10 years ago or today? (Ignore the fact that the subtitle implies you’ll be looking at percent change over time but the bars show levels.)

Bar and line chart titled 'China's Fiscal Spending Not a Boost to Growth Yet,' subtitled 'Spending was meant to rise 3% this year, but is down 1% so far.' Stacked bars from 2017 to 2024 show actual spending through September in pink and projected rest-of-year spending in yellow, against a black line showing the annual budget forecast, rising from about 26 trillion yuan in 2017 to roughly 41 trillion yuan in 2024. The 2024 bar shows only the pink actual-spending portion, which falls short of the budget line, illustrating that spending is lagging behind the planned increase. Source: China's Ministry of Finance. Bloomberg.

It’s also a curious decision to place the labels on the right side of the graph because I wonder whether it encourages our eye to move quickly from the title—which evidence suggests people do read—to the other side of the chart. Does that mean we miss the left side of the chart, which helps set a baseline for where the data started? Or is that okay because we are encouraging them to get to where the data are today

You will sometimes see creators place y-axis labels on both sides of a chart, as in the example below showing real hourly earnings. It’s worth noting this isn’t a true dual-axis chart in the problematic sense where two different variables plotted on two different scales; here, it’s the same scale mirrored on both sides, which makes it a cleaner test case for the real question: Does repeating the labels help orient the reader, or does it just add clutter? Personally, I’m not sold that placing labels on both sides is particularly helpful, though I can see the appeal on a chart with this many lines and space between them.

Line chart titled 'Real Hourly Earnings by Education Status, Men Aged 25–54, 1965–2016.' Five lines track average hourly wages in 2017 dollars from 1965 to 2015 by education level. Advanced degree holders rise from about $23 to $38, college grads from about $23 to $30, with both showing sharp growth after 1995. Some college holders hover near $20, fluctuating between $17 and $24. High school graduates decline from about $18 to $17.50, dipping to $17 in the 1990s before a slight recovery. High school dropouts fall from about $14.50 to $14, dropping as low as $13 in the early 1990s. The chart shows a widening gap between higher and lower education groups over the five decades.

As I continue my thinking on this, I think I’ve landed on a few considerations:

  • Placing the yaxis labels on the right side of the chart is especially helpful if the most recent data is where you want to direct people’s attention.
  • Depending on the patterns in the data, it may make adding direct labels to the chart more difficult.
  • This might be an easy way to make charts designed for left-to-right readers more accessible for right-to-left readers.

Want to learn more?

What do you think? Right-sided labels: Fan or Foe?

Related Posts