The electoral college map and why visualization matters
Much of this piece will be obvious to people who spend time thinking about things like the electoral college map. But the importance of the electoral college — its implications for representative democracy, federalism, and the even the factors political parties weigh when nominating their candidates for the presidency —is far from obvious to more casual observers, and perhaps for many voters. That the tiniest of swings in particular parts of particular states can determine who wins and who loses in the race for the White House is perhaps even less obvious.
As a student of American politics, I found studying the electoral college map and toying with alternative scenarios endless fun. Going back through history to see how the map changed — particularly the significant changes in the map from the 1960s through to the now expected blue west coast and north-east, red interior. Seeing how presidents like Johnson, Reagan, and Clinton built their formidable victories (and how different those victories looked on the map) is an exercise of fascination.
The electoral college and its relationship with the popular vote
While most American adults are given the right to vote for president, the reality is that those votes are not collated together and counted to see who has earned the most. Elections for president are more like a series of 51 elections, held by each state (plus DC), rather than one national election. It’s the outcome of each of those elections within the series that determines who wins the presidency. To boil it down:
- The candidate who earns the most votes in a state wins that state’s electoral votes (with two exceptions: Maine and Nebraska aren’t ‘winner take all’).
- Each state is ‘worth’ a certain number of electoral votes, roughly corresponding to population. The minimum number is 3, which smaller states like Wyoming, Delaware, and DC have, right up to 55 for California, the most populous state.
- In total, there are 538 electoral votes. To win the presidency, a candidate needs to win more than half of those 538 electoral votes, so 270 or more.
Building a path to 270 electoral votes is a matter of winning a combination of states whose electoral votes together reach that magic number, meaning the difference between winning and losing often comes down to a relatively small number of votes in key battleground states.
In 2016, Donald Trump received around 3 million fewer votes than Hillary Clinton, but won the election. Why? Because he earned very narrow victories in three key states: Pennsylvania, Michigan, and Wisconsin, which together are worth 46 electoral votes. Clinton won by large margins in big states like California with its 55 electoral votes, and New York with its 29, but it wouldn’t have mattered whether she won those states by 1 vote or 1 million votes. In the case of California, Clinton won 4 million more votes than Trump, but she would have received the same 55 electoral votes from California had she won the state by a much narrower margin. The implication is that there is often a misalignment between the national popular vote —the actual total votes each candidate gets across the country — and the electoral college vote.
One of the most powerful ways to understand this dynamic is visualization. Electoral cartographers produce stunning representations of things like concentrations of particular voting behaviours, and the distribution of votes by population rather than geographic state boundaries.
This first map shows the outcome of the 2016 election. Visually, red dominates. While Donald Trump won with a reasonably significant electoral college margin — 306 to 232 —a glance at the map suggests a landslide.
It’s easy to see why: the red states of Idaho, Montana, Wyoming, North and South Dakota, and Nebraska in the upper left of the map together are home to around 7 million Americans, only 36% of the 19 and a half million New Yorkers, a small slice of blue in the north east.
The second map rescales states in accordance with their population. Suddenly, the visual cue is of a significantly closer election.
But even then, the map remains a fairly blunt instrument. The swathes of blue on the west coast and north east, the unbroken belt of red running from the Great Lakes to the Gulf of Mexico and across the south, depict a uniformity that belies the reality of far less homogenous regions.
The 2016 presidential election, as well as powerfully illustrating the misalignment between the national popular vote and the electoral college, was exceptionally close in a number of states. Donald Trump won Michigan by just over 10,000 votes, Wisconsin by a shade under 23,000. Hillary Clinton won New Hampshire by less than 3000 votes.
In looking at the 2016 map and thinking about the 2020 presidential election, I found myself wanting to see exactly what it would take for Joe Biden to win Pennsylvania, or Michigan or Wisconsin; for Donald Trump to win Nevada or New Hampshire. If 3% of voters from 2016 changed their vote this time around in one direction or the other, how would the map look?
And what about turnout? Some elections, and some candidates, excite voters more than others. Whether voters switch sides is no more important than whether rusted on voters turn up in the same numbers they did last time.
Wanting to see the answers to these questions visually, I worked on an electoral college map with the voting results data of two previous elections (2012 and 2016) built into it. While the data can be used in any number of ways, I focused on the two key and most commonly used indicators of voting behavior — swings and turnout.
When the data is applied through tools we’re used to — in this case, simple sliders — the extreme closeness of elections (or at least, parts of them) becomes apparent. Being able to visualize the effect of a mere 0.5% swing toward Joe Biden in 2020 over the 2016 results (giving him 278 electoral votes to Donald Trump’s 260). Users can see the effect of changes nationally, regionally, or state by state, gaining insight on the fundamental question of ‘what would it take’ for a certain outcome to happen, or projecting outcomes based on polling, for example.
Exit poll data, which contains information on how demographic groups voted, is included as well, and like other (and excellent) electoral map tools available, users can click on states to automatically flip them from one party to the other, and quickly build pathways to 270 electoral votes.
I look forward to contributing to the wealth of discussion on the electoral college and the election, and hope others find use in visualizing what it takes for the electoral college map to shift its hues again.
J A Saunders runs the data driven electoral college map website www.thepolivision.com
