A cartogram is a map that deliberately distorts the size or shape of geographic areas to reflect a data variable like population, GDP, or disease rates instead of actual land area. The core purpose is to replace physical geography with data-driven geography, so that what you see on the map directly represents the thing you care about measuring. A massive but sparsely populated region like Montana shrinks, while a small but densely packed state like New Jersey swells.
How Cartograms Work
On a standard map, each country or state appears at its true geographic size. That works fine for navigation but creates a problem when you’re trying to visualize data tied to people, money, or resources. Canada looks enormous and Singapore looks like a pinprick, even though Singapore might outpace Canada on the variable you’re mapping. A cartogram solves this by scaling each area proportionally to the data it represents. If you’re mapping global CO₂ emissions, China and the United States balloon while vast but low-emitting regions in central Africa contract.
The mathematical process behind this involves transforming the map so that each region’s new size matches its share of the data variable. Early computer-generated cartograms, dating to the late 1960s, divided maps into small cells that were independently expanded or shrunk, then repeated the process until the sizes stabilized. Modern approaches, like the widely cited diffusion method developed by physicists Michael Gastner and Mark Newman, treat the data as a density that “flows” across the map the way heat spreads through a material. The result is a smoother, more recognizable distortion that preserves the general arrangement of regions while resizing them.
Correcting the Landmass Bias in Elections
One of the most popular uses of cartograms is visualizing election results. A standard red-and-blue map of U.S. presidential elections gives a misleading impression because geographic area is not equivalent to votes. Rural states with small populations dominate the visual landscape, making one party’s support appear far more widespread than it actually is. Scaling state sizes by population or electoral votes corrects this distortion, showing actual levels of political support rather than the accident of how much empty land a state contains.
This application has a surprisingly long history. One of the earliest examples in the American press appeared in the Washington Post in 1929, where state areas were scaled by population and federal tax payments to argue that giving every state equal voting weight on tariff measures was unfair. The technique gained mainstream attention again after the 2004 presidential election, when population-scaled cartograms circulated widely online. The Washington Post described the resulting images as making “America look like a flabby cartoon character stretched into blue and red spandex.” The visual impact was immediate: suddenly the Northeast corridor and California carried the visual weight their voter populations warranted, while the Mountain West receded.
Public Health and Disease Mapping
Epidemiologists use cartograms to reveal disease patterns that standard maps obscure. If you plot lung cancer cases on a regular map of New York State, you’ll see clusters in geographically large rural counties that may only reflect their size on the map, not an actual concentration of illness. Redrawing the map so each county’s size reflects its population puts the case counts in proper context.
Researchers have used this approach to visualize global drowning incidence, deaths from heart disease, HIV incidence in Japan, obesity rates in Canada, and lung cancer cases across New York. In one study, sheep scrapie occurrence in Great Britain was mapped on a cartogram scaled to the sheep population by county, making it possible to see whether outbreaks were concentrated relative to the number of animals at risk rather than relative to geographic space. A framework for analyzing leukemia incidence rates across California counties from 2008 to 2013 used cartograms to test whether apparent disease clusters were statistically meaningful or just artifacts of uneven population distribution.
The key insight in all these cases is the same: when the thing you’re counting (people getting sick, animals being tested) is unevenly distributed across space, a regular map will mislead you. A cartogram redraws the canvas to match reality.
What Types of Data Work Best
Cartograms are most effective when there’s a large mismatch between a region’s physical size and the variable being mapped. Population is the classic example because the difference between geographic area and population is dramatic in most countries. Russia covers more land than any other nation but has a smaller population than Bangladesh, which is roughly the size of Iowa. A population cartogram makes that contrast visceral in a way no bar chart can.
Beyond population, cartograms commonly visualize GDP, carbon emissions, internet users, military spending, refugee flows, and research output. Any count or total that accumulates unevenly across geography is a good candidate. Rate-based data (like income per capita or life expectancy) can also work, though the visual effect is less striking because rates don’t compound with population size the way totals do. The technique is less useful for data that correlates closely with land area, like agricultural acreage or forest cover, since the cartogram would barely differ from a standard map.
Cartograms vs. Choropleth Maps
The most common alternative to a cartogram is a choropleth map, which keeps geographic shapes intact but uses color shading to represent data values. Choropleths are easier to read because the familiar outlines of countries or states remain unchanged. But they have a built-in flaw: large regions dominate your visual attention regardless of their data values. A lightly shaded Siberia still commands more of the map than a darkly shaded Belgium, which can make small but significant regions nearly invisible.
Cartograms flip this trade-off. They give regions visual weight proportional to their data, making it easy to compare the actual importance of different areas at a glance. The cost is that distorted shapes can be harder to recognize. Readers may struggle to identify which bloated or shrunken region is which, especially in maps with many small units. This is why cartogram designers often label regions directly or include a standard map alongside the cartogram for reference. The best choice between the two depends on whether your priority is geographic familiarity or data accuracy.
Common Types of Cartograms
Not all cartograms distort geography the same way. The most common types each handle the trade-off between recognizability and precision differently.
- Contiguous cartograms keep all borders connected, stretching and squeezing regions while maintaining their neighbors. These look the most like a real map but can become hard to read when the data forces extreme distortions.
- Non-contiguous cartograms separate regions from each other, resizing each one independently. The shapes stay recognizable because they aren’t warped by their neighbors, but the gaps between them sacrifice the sense of geographic continuity.
- Dorling cartograms replace each region with a circle (or other simple shape) sized to the data value. Geographic accuracy disappears almost entirely, but the data comparison becomes extremely clean. These are popular in data journalism for their visual clarity.
- Grid or tile cartograms represent each unit as an equal-sized square or hexagon arranged in roughly the correct geographic position. Every unit gets the same visual weight, which is useful when you want to show categorical data (like which party won each state) without any size distortion at all.
Why Cartograms Are Hard to Read
The biggest limitation of cartograms is cognitive load. People rely on the familiar shapes of countries and states to orient themselves on a map. When those shapes are warped, readers spend mental energy just figuring out what they’re looking at before they can absorb the data. This is especially true for contiguous cartograms where extreme distortion makes regions unrecognizable.
Designers mitigate this by choosing the right type of cartogram for the audience. A Dorling cartogram with labeled circles works well in a news article where readers need a quick takeaway. A contiguous cartogram works better for an audience that already knows the geography and wants to see how regions compare in context. Including a small reference map in the corner helps readers match distorted shapes to real locations. Color coding by category or value also helps, since readers can identify groups of regions even when individual shapes are ambiguous.