The Mercator projection preserves shapes and directions but grossly exaggerates the size of land near the poles; the Gall–Peters projection preserves true relative area but badly distorts shapes. Neither is "the correct map." They are opposite trade-offs, and the mathematics guarantees that every flat world map must make one — you cannot flatten a sphere without breaking something.
The Mercator–Peters argument is the most public dispute in the history of cartography. It has run through academic journals, school board meetings and prime-time television, and it is unusual among map debates in that both sides are partly right. Understanding why requires separating what each projection does mathematically from what people have claimed about it politically.
What the Mercator projection actually does
Gerardus Mercator published his world map in 1569 to solve one specific problem: sea navigation. On his projection, a line of constant compass bearing — a rhumb line — appears as a straight line on the map. A navigator could lay a ruler between two ports, read off a single bearing, and hold that heading. No other projection in general use does this, and it is why the Mercator became and remains the standard for nautical charts.
The property that makes this work is conformality: the projection preserves angles, and therefore local shapes, at every point. The price is paid in area. To keep angles correct as the meridians are pulled apart into parallel vertical lines, the map must stretch the north–south direction by the same factor. That factor grows rapidly with latitude, and the resulting area exaggeration grows as roughly the square of it. At 60° north, land appears about four times too large. At 80°, roughly thirty times. At the poles, the factor becomes infinite, which is why the Mercator cannot show the poles at all and every Mercator world map is simply cut off somewhere in the Arctic and Antarctic.
The consequences are the familiar ones. Greenland looks about the size of Africa. In reality Africa covers roughly 30 million square kilometres and Greenland roughly 2.2 million — Africa is about fourteen times larger. Alaska looks comparable to Brazil, which is more than four times its size. Europe and North America appear far more prominent than their true share of land.
What the Gall–Peters projection actually does
The Gall–Peters is a cylindrical equal-area projection with standard parallels at 45°. Equal-area means exactly what it says: any two regions that cover the same area on the globe cover the same area on the map, everywhere, without exception. Africa appears fourteen times the size of Greenland because it is fourteen times the size of Greenland.
The mathematics is not new. James Gall, a Scottish clergyman, described the projection in 1855. Arno Peters — a German historian and film-maker, not a trained cartographer — presented it independently in 1973 as "the Peters projection," framed explicitly as a corrective to what he called the Eurocentric bias of the Mercator. The name Gall–Peters is now the standard credit for both.
The trade-off runs in the opposite direction to the Mercator's. Because area must be conserved while the meridians are held parallel, whatever is stretched in one direction must be squeezed in the other. Landmasses near the equator — Africa, South America, Indonesia — are pulled vertically into elongated shapes. Landmasses at high latitudes are flattened horizontally. Arthur Robinson, one of the twentieth century's most influential cartographers, memorably described the result as looking like wet, ragged long winter underwear hung out to dry.
The distortions side by side
- Mercator preserves: angles, local shape, compass bearings. Destroys: area, increasingly toward the poles.
- Gall–Peters preserves: area, exactly and everywhere. Destroys: shape, most severely near the equator and the poles.
- Both destroy: distance and, for practical purposes, the ability to compare directions across the whole map.
- Both are rectangular, which forces the north–south and east–west scales apart at every latitude except the standard parallels.
This is not a failure of imagination on anyone's part. It is a theorem. A sphere and a plane have different intrinsic curvature, so no flat map can represent a globe without distortion of some kind — and in particular, no projection can be both conformal and equal-area at the same time. Cartographers quantify what is lost using Tissot's indicatrix, a set of small circles drawn on the globe and mapped onto the projection: on a Mercator they stay circular but swell enormously toward the poles, while on a Gall–Peters they keep constant area but deform into ellipses.
The political argument, and the pushback
Peters did not present his map as a technical improvement. He presented it as a moral one, arguing that centuries of Mercator wall maps had trained Western students to see the tropics — Africa, South Asia, South America — as smaller and less significant than they are, and Europe and North America as larger and more central.
The empirical part of that claim is hard to dispute. Mercator maps were used for decades in classrooms where navigation was irrelevant and area comparison was the whole point, and generations did come away believing Greenland rivals Africa. The debate reached a mass audience when a 2001 episode of The West Wing staged a fictional lobbying group pitching the Peters map to the White House, and it resurfaced in 2017 when Boston Public Schools became the first large US district to switch its classroom maps to Gall–Peters.
The pushback came from cartographers, and it was pointed. Mercator designed his projection in 1569 for navigation, not ideology; attributing colonial intent to a sixteenth-century mathematical construction confuses a tool with its later misuse. Critics also noted that Peters overstated his originality — the projection was Gall's, more than a century earlier — and made technical claims about distance and directional fidelity that simply were not true of his map.
What professional cartographers actually recommend
The cartographic consensus is that the debate is a false choice, and the profession said so formally. In 1989 several major North American geographic organisations passed a joint resolution urging that rectangular world maps be abandoned for general-purpose use — a statement aimed at the Mercator and the Gall–Peters alike, because the rectangular frame is what forces the worst of both distortions.
If the goal is honest area comparison, there are equal-area projections that achieve it with far less shape damage than Gall–Peters, because they do not insist on a rectangle:
- Mollweide — an elliptical equal-area projection, in use since 1805, with curved meridians that greatly reduce the stretching.
- Eckert IV — equal-area with a pole line rather than a pole point, giving high latitudes more reasonable proportions.
- Equal Earth — introduced in 2018 specifically as a visually appealing equal-area alternative, and now widely adopted for thematic world maps.
- Hobo–Dyer — a cylindrical equal-area projection like Gall–Peters, but with standard parallels around 37.5°, which shares the area accuracy while looking noticeably less distorted.
Mercator vs Peters vs the compromise projections
There is a third option the debate tends to obscure: projections that deliberately preserve nothing perfectly in order to distort everything mildly. These are the compromise projections, and they are what most general-reference world maps actually use.
The Robinson projection, devised in 1963, was built by eye rather than by formula — Arthur Robinson adjusted the curves until the world simply looked right — and the National Geographic Society used it for its reference world maps from 1988 to 1998. The Winkel Tripel, published in 1921, takes its name from the German Tripel, "triple," because it aims to minimise three kinds of distortion at once: area, direction and distance. National Geographic replaced Robinson with Winkel Tripel in 1998, and it remains the most common choice for authoritative world maps today.
Set against each other, the practical guidance is straightforward:
- Navigating, or looking at a city? Mercator. Its distortions are imperceptible at local scale, which is why Web Mercator underpins nearly every online slippy map.
- Comparing how much land countries or biomes cover? An equal-area projection — but choose Equal Earth, Mollweide or Eckert IV over Gall–Peters.
- A general-purpose wall map? Winkel Tripel or Robinson.
- Teaching that maps are arguments? Show several, side by side. That is the lesson the whole controversy actually delivered.
Seen this way, the Mercator–Peters fight produced something more valuable than a winner. It made a large public audience aware that maps can mislead, that the choice of projection is an editorial decision, and that "accurate" is not a property a world map can simply possess — it depends entirely on what you are trying to measure.
Frequently asked questions
Which is more accurate, Mercator or Peters?
Neither, in the abstract. Mercator is accurate for angles and shapes; Gall–Peters is accurate for area. Ask what you need the map to get right, and the answer follows.
Why does Google Maps still use Mercator?
Because at street and city zoom levels the area distortion is negligible, while the shape and angle fidelity matters a great deal for navigation. Google Earth's globe view avoids the problem entirely by not flattening the world at all.
Is the Peters projection the same as the Gall projection?
Mathematically, yes. James Gall described it in 1855; Arno Peters arrived at it independently in 1973. "Gall–Peters" is the standard name.
Is there a projection with no distortion?
Not on a flat sheet. Only a globe represents shape, area, distance and direction correctly at the same time. Every flat map chooses which errors to accept.