🟣🔵 Bivariate maps: Yay or Nay?

👋 Hi there,

By now, you probably know how much I love data visualization and exploring new chart types.

And yet, there is one type I have barely mentioned over the years: bivariate maps.

I have always had mixed feelings about them. So today, let’s finally talk about it.

What is a bivariate map?

A bivariate map is a variation of a choropleth map that displays two variables at the same time within the same geographic areas.

Instead of coloring regions according to a single metric, it combines two variables into a shared color scale. Each color represents a specific combination of values, usually shown through a two dimensional legend.

It is powerful! You can reveal relationships between variables directly on the map. Correlations, contrasts, clusters. All in one view.

I should definitely add a new category on my Dataviz Inspiration project.

But… is it actually good?

That is where things get interesting.

I shared another example on LinkedIn and the discussion that followed was fascinating. Opinions were strongly divided. And I realized I am not alone in struggling with these maps.

Yes, displaying two variables in the same geographic space is efficient. It packs a lot of information into a single visual.

But the cognitive load can be intense.

Personally, I find myself constantly going back to the legend. Over and over again. Trying to decode what each color actually represents before understanding the story.

More often than not, two separate choropleth maps placed side by side feel clearer and easier to interpret.

I wish I could give you a definitive answer here. Use them. Avoid them. Follow this rule.

But that is also what makes this field so fascinating. There is rarely one universal truth everyone agrees on.

Can interactivity make them better?

Now this is where things get exciting.

The example I shared included a small interactive feature that completely changed the experience for me.

When hovering over the legend, the corresponding areas on the map are highlighted.

Such a simple idea. Yet incredibly effective.

It helped me immediately understand patterns that felt almost unreadable in the static version.

More broadly, I believe interactive legends are becoming an essential design pattern. When used well, they reduce cognitive effort and guide attention exactly where it needs to go.

Conclusion & Next step

Even after 12 years working in this field, there are still topics I feel uncertain about.

And honestly, I like that!

This interactive example reinforced something I strongly believe: as tools evolve and AI enables more bespoke and interactive visualizations, the quality and clarity of our work can dramatically improve.

Static charts force compromises. Interactive ones allow exploration if interactivity is used properly.

Now, I want to push this idea further. Imagine hovering over one of the two variable names and temporarily switching to a standard choropleth of that variable alone. That would make comparison far more intuitive!

I will experiment with it and share the results with you soon.

For now, I need to get back to writing lessons for my D3.js students, who are currently learning how to build exactly this kind of interactive crazyness!

See you next week,

Yan