The Peters Projection

The Peters projection is the most famous of the equal-area cylindrical projections — and the most controversial map in the history of cartography. The /learn/what-is-a-map-projection pillar covers Gauss's Theorema Egregium and the impossibility of preserving every property at once; the /learn/mercator-projection support covers the conformal projection Peters argued against. This article covers the geometry of Peters, the 1855 origin and 1973 reintroduction, the 1980s “map wars”, and the modern view that places Peters in the broader family of equal-area projections covered in /learn/equal-area-projections.

What the projection does

The Peters projection is cylindrical: meridians are equally spaced vertical lines, parallels are horizontal lines. The horizontal scale is set by the standard parallels at ±45°; at those two latitudes, both horizontal and vertical scales match Earth's true scale. Everywhere else, the scales differ — but the product of the two scales is held constant, so the area on the map equals the area on Earth.

On a sphere of radius R, the Peters projection coordinates are:

x = R · (λ − λ₀) · cos 45°
y = R · sin φ / cos 45°

where λ is longitude, λ₀ is the central meridian, and φ is latitude. The ratio of the map's width to its height is 2π · cos 45° : 2 / cos 45° = π : 1 — taller than the equatorial aspect ratio, shorter than the polar one.

The area-preserving property is exact in the same way Mercator's angle-preserving property is exact. Every region's area on the map equals its true Earth-surface area in proportion. Africa is shown at its proper 30.4 million km², about 14 times the size of Greenland at 2.16 million km² — a ratio Mercator visually inverts.

What the projection does badly

Shape distortion is the price. Per Snyder's Map Projections, shape distortion in Peters is severe and asymmetric:

• Near the equator, parallels are stretched too far horizontally relative to the vertical compression; landmasses appear elongated east-west.
• Near the poles, the same forces invert; high-latitude regions appear flattened vertically.
• The standard parallels at ±45° are the only latitudes where the scale is correct in both dimensions.

Visually, equatorial Africa and South America look stretched and thin; North America and Eurasia look squashed and flattened. The effect is striking on first viewing — readers used to Mercator's “normal” world map find Peters disorienting.

The shape distortion is no worse than Mercator's area distortion in any rigorous sense; both projections preserve one property completely while sacrificing another. The /learn/equal-area-projections support covers alternatives — Mollweide, Eckert IV, Hammer — that preserve area while distributing shape distortion more pleasantly.

James Gall and the 1855 origin

James Gall, a Scottish clergyman and amateur cartographer, described the projection in 1855 in his paper “On Improved Monographic Projections of the World”, presented at the British Association for the Advancement of Science. Gall called it “orthographic” (unhelpfully, since “orthographic” was already used for a different projection family) and proposed it as one of three projections he had developed; the other two were the Gall stereographic and the Gall isographic.

The 1855 paper was published in The Scottish Geographical Magazine but the projection saw little use through the rest of the 19th and early 20th centuries. Gall's prior work was forgotten by the 1960s — even most professional cartographers were unaware of it when Peters introduced his version in 1973.

Arno Peters and the 1973 reintroduction

Arno Peters, a German historian and filmmaker, presented his projection at a Bonn press conference on 5 May 1973. He called it simply the “Peters Projection” and made no reference to Gall's prior work. The presentation accompanied an explicit political argument:

• The Mercator projection, then the standard world-map projection in Western atlases, made the northern temperate countries (Europe, North America, Russia) appear larger than they were while making equatorial countries (Africa, Latin America, Southeast Asia) appear smaller.
• This was a colonial-era artefact that distorted Western perceptions of the developing world.
• His projection corrected the distortion by preserving area.

Peters was an effective publicist. His map was adopted by UNESCO, the United Nations Development Programme, the World Council of Churches, several African governments, Oxfam, and dozens of NGOs. For nearly two decades, Peters' map was the default development-aid world-map and appeared in classrooms across Britain, Germany, and much of the global South.

The pure cartographic content was unoriginal — Gall's 1855 projection was the same — but the political framing was new and powerful.

The 1980s map wars

The American cartographic community responded slowly but firmly. Arthur H. Robinson, who had himself designed the /learn/robinson-projection for National Geographic in 1963, published “Arno Peters and His New Cartography” in The American Cartographer in 1985, calling the projection “somewhat reminiscent of wet, ragged long winter underwear hung out to dry on the Arctic Circle”. The journal devoted a substantial portion of that year's issues to the controversy, with several articles criticising the projection.

The most consequential response came in February 1989, when seven cartographic societies — the American Cartographic Association, the American Geographical Society, the Association of American Geographers, the Canadian Cartographic Association, the National Council for Geographic Education, the National Geographic Society, and the Special Libraries Association Geography and Map Division — issued a joint Resolution on Rectangular World Maps. The resolution called on book publishers, the media, and government agencies to “cease using rectangular world maps for general purposes or artistic displays”, including both Peters and Mercator, recommending pseudocylindrical and other compromise projections instead.

The criticism was largely technical: any cylindrical equal-area projection has the same area-preserving property as Peters; the projection is one of an infinite family (the Lambert cylindrical equal-area projection at standard parallels of 0° is the limit case). Peters' specific choice of 45° standard parallels was no more “correct” than any other choice. The political argument about Eurocentrism applied equally to many projection families, not just to Mercator.

Peters defended his projection in books, interviews, and lectures until his death in 2002. His followers — the “Peters Society” and various NGO collaborators — continued promotion afterward.

The West Wing moment

The projection received a second wave of attention from a different source: the West Wing television series. In Season 2, Episode 16 (“Somebody's Going to Emergency, Somebody's Going to Jail”, aired February 2001), the fictional “Organisation of Cartographers for Social Equality” visits the White House to argue for replacing Mercator with Peters in US public schools. The episode portrayed the Peters argument sympathetically and brought the controversy to a much wider audience than it had previously reached.

Several US school districts subsequently adopted Peters. Boston Public Schools in 2017 switched its standard classroom world map to Gall-Peters, citing the West Wing-style argument and Peters's political framing.

The cylindrical equal-area family

Peters is mathematically a special case of the broader Lambert cylindrical equal-area projection family, parameterised by a single standard parallel φ₀:

x = R · (λ − λ₀) · cos φ₀
y = R · sin φ / cos φ₀

Varying φ₀ produces every member of the family. Each has different shape trade-offs:

| Projection | Standard parallels | Aspect ratio (W:H) | Visual signature | |---|---|---|---| | Lambert cylindrical equal-area | 0° (equator) | π : 1 | Severe vertical stretching toward poles | | Behrmann (1910) | ±30° | 2.36 : 1 | Compromise; popular in textbooks | | Smyth equal-surface (1870) | ±37.5° | 2.0 : 1 | Roughly square countries near standard parallels | | Hobo-Dyer (2002) | ±37.5° | 2.0 : 1 | Same as Smyth; modern rebranding | | Gall-Peters | ±45° | π/2 : 1 ≈ 1.57 : 1 | Tall map; severe equatorial stretching | | Tobler hyperelliptical (1973) | ±55° | Tall | Polar regions correct, equatorial bad |

The Peters projection's ±45° standard parallels are a deliberate choice — Peters wanted a projection that emphasised equatorial regions visually. The choice is arbitrary in the mathematical sense; every member of the family is equally area-preserving.

A worked example: Greenland and Africa

The classic Peters-vs-Mercator comparison concerns Greenland and Africa. The actual areas, from USGS and CIA World Factbook data:

• Greenland: 2.166 million km²
• Africa: 30.37 million km² (the continent, not just the Sahara)

Africa is about 14.0 times the area of Greenland. On the Mercator projection, however, Greenland appears about the same size as Africa — both project to roughly equal vertical extent because Mercator stretches everything in the high latitudes. On the Peters projection the areas are correct: Africa appears 14 times the area of Greenland.

The visual contrast is striking. A Peters map shows Africa as a massive, dominant feature of the world map; a Mercator map shows Africa as merely one of several large landmasses, comparable to Greenland in apparent size. The political effect Peters argued about is real — even though the choice of which property to preserve is inherently a trade-off rather than a moral judgement.

Modern view

Contemporary cartography places Peters in context:

• It is one of an infinite family of cylindrical equal-area projections. The Behrmann (1910) at standard parallels ±30°, the Hobo-Dyer (2002) at ±37.5°, and the Smyth equal-surface (1870) at ±37.5° are all in the same family; they differ in the trade-off between vertical and horizontal stretching.
• It preserves area but distorts shape; a fully balanced compromise is offered by the Robinson, Winkel Tripel, or Natural Earth projections covered in their respective supports.
• Web mapping uses Web Mercator (a spherical Mercator) for tile-based zoom-and-pan; equal-area projections are typically reserved for static thematic maps where area comparison is important.
• The political argument retains some validity at a meta level (cartographic choices have rhetorical consequences) but the technical claim that Peters is uniquely “correct” is not sustained.

The Peters projection is now historically significant rather than operationally important. Almost no professional cartographer uses it as a first choice; it survives in NGO outreach materials and in specific classroom contexts where the political framing is the intended message.

Sources

• Snyder, Map Projections — A Working Manual (USGS Prof. Paper 1395) — equations and mathematical context.
• Snyder, Flattening the Earth (University of Chicago Press, 1993) — the definitive history of map projections including the Peters controversy.
• Robinson, “Arno Peters and his New Cartography” (The American Cartographer, 1985) — the cartographic-community response.
• Resolution on Rectangular World Maps (1989) — the seven-society joint statement.

For closely related topics, see /learn/mercator-projection for the projection Peters argued against, /learn/equal-area-projections for the family Peters belongs to, and /learn/what-is-a-map-projection for the impossibility theorem behind the trade-offs.