Look at a map of the 2016 US presidential election and it appears to be overwhelmingly red. Vast stretches of the country, from the Great Plains to the deep South to rural New England, colored in a single uniform shade. Now look at the vote totals: Donald Trump received 46.1% of the popular vote, Hillary Clinton received 48.2%. A map that looks nearly monochromatic turns out to represent a race that was almost perfectly split. Something is wrong with this picture — and understanding what is wrong tells you something important about how to read any election map, not just American ones.
Maps show land, not people
The fundamental problem is simple. A choropleth map — the type that colors regions based on which party won them — assigns visual space proportionally to geographic area, not to population. A sparsely populated rural county that covers 5,000 square kilometers and has 20,000 voters gets far more space on the map than a dense urban district covering 200 square kilometers with 500,000 voters.
In the United States, this distortion is extreme because population is concentrated in a small number of metropolitan areas. Wyoming covers 253,000 square kilometers and cast roughly 255,000 presidential votes in 2020. Los Angeles County covers 10,500 square kilometers and cast over 3.3 million votes. Wyoming is 24 times larger on the map; Los Angeles County cast 13 times as many votes. A map-reader looking at area naturally concludes that Wyoming matters more. A voter-count reader sees the opposite.
The same problem appears in every country
This is not uniquely American. The same geographic distortion appears in any country where voting patterns differ between densely and sparsely populated regions — which is to say, almost every country.
The UK's 2019 general election illustrates it vividly in a different direction. London, which cast over 2.5 million votes and is one of the most politically distinct regions in the country, occupies a tiny sliver of the UK map. The vast agricultural expanses of eastern England and the Scottish Highlands dominate visually, despite contributing relatively few votes. A map of 2019 constituency results looks almost entirely Conservative blue — yet Labour won 32.2% of the national vote and the Conservatives only 43.6%.
Hungary's 2022 parliamentary election provides an example ElectioMap covers directly. Fidesz dominated the geographic map of Hungary, winning most constituencies outside Budapest. But a glance at the map exaggerates the margin significantly. Budapest and its suburbs voted strongly for the opposition, and while those areas are small on the map, they account for a large share of the total electorate.
Poland's elections tell a similarly striking spatial story. The 2023 parliamentary election produced what commentators called "two Polands" — eastern and rural areas voting heavily for PiS, western and urban areas voting for the opposition coalition. The map suggests a country divided roughly in half by area, but the voter division was different: the opposition coalition together collected more votes than PiS, even though PiS won more individual constituencies.
What election maps do and don't show
None of this means election maps are useless. They are genuinely informative — just not about what people often assume they show.
A geographic choropleth map does tell you:
- Where, geographically, each party's support is concentrated
- Whether support is uniform or regionally clustered
- The urban/rural pattern of the vote
- Which regions swung between elections
A geographic choropleth map does not tell you:
- How many votes each party actually received
- How close or decisive the overall result was
- Which side "won" the country in any meaningful sense beyond seats
- What the electorate as a whole wanted
Better ways to visualize vote data
Cartographers and data journalists have developed several approaches that address the area-vs-population problem. Cartograms resize each region proportionally to its population or vote count rather than its geographic area, producing maps that look distorted but represent voter weight accurately. Dot-density maps place one dot for every fixed number of votes, letting population density speak for itself. Graduated color scales — showing the margin within each region rather than just the winner — add another layer of honest information.
On ElectioMap, we use choropleth shading with color intensity scaled to the winning margin rather than simply coloring each region in the winner's full color. A region won 51-49 looks noticeably different from one won 75-25. This does not solve the area problem, but it at least prevents the map from implying that every district was a landslide for one side. We always display the full vote-share breakdown alongside the map, so the numbers are visible even when the geography obscures them.
The best habit when reading any election map is to look at the numbers first and the map second. The numbers tell you what happened; the map tells you where. Both matter, but they answer different questions — and only one of them tells you who won.