Monday, November 2, 2020

What Really Happened in the Electoral College, II

In a previous post, I analyzed the reasons behind the wide divergence between the Electoral College and national popular vote in 2016, finding that although the Electoral College had a slight bias in favor of Obama in 2012, Trump was able to improve his margins in the crucial Great Lakes states while Hillary Clinton's improvements mostly came in California and Texas, neither of which switched any electoral votes. A further comparison of the 2012 and 2016 elections, and the shifts in the states between them, sheds light on the real function of the Electoral College in our republic. 

So, the bias in the Electoral College toward Republicans that proved decisive in 2016 was not present in 2012. Opponents of the Electoral College have a point in noting that it has a bias in favor of small states, since each state is guaranteed two electors for each of its senators. In both recent elections where the Electoral College and popular vote conflicted, the Republican carried thirty states while the Democrat carried twenty and the District of Columbia. If the Electoral College only allocated electors based on House representation, the Republican’s total would drop by sixty in both cases, while the Democrat’s would drop by forty-two. This would be enough to reverse the 2000 election results (from a 271-267 Bush victory to a 225-211 Gore one) but not 2016. Donald Trump would have won by a narrower margin (246 to 190, instead of 306 to 232), but he still would have won the Electoral College despite losing the popular vote by a wider margin than Bush.

If the Electoral College’s bias toward small states alone does not account for Trump’s victory, what else does? In 2012, President Obama got sixty electoral votes from states he carried by less than five percent of the vote. He received a further eighty-two electoral votes from states (and the second congressional district in Maine) that he carried by between five and ten percent of the vote, for a total of 142 electoral votes, or just over forty percent of his total of 332 electoral votes. By contrast, Mitt Romney carried only one state (North Carolina) by less than five percent, and only three (Missouri, Georgia, and Arizona, plus one electoral vote from a congressional district in Nebraska) by a margin of between five and ten percent. Only fifty-three of Romney’s 206 electoral votes came from these states. Put another way, Obama won the electoral votes of states decided by ten percent or more fairly narrowly (190 to 153), but won a near three-to-one margin among states decided by less than ten percent (142 to 53). Where Romney won, he tended to win big; Obama carried most of the states that were decided by narrower margins.

    




           In 2016, this trend reversed between the parties and became even more pronounced. Among states and districts that were decided by ten percent of the vote or more, Hillary Clinton trounced Donald Trump, 183 electoral votes to 126. However, Hillary Clinton only carried two states by between five and ten percent, and five states by less than five percent, getting only forty-nine additional votes. Donald Trump, however, received more than half his electoral votes- 180 out of 306- from states he carried by less than ten percent of the vote.  Trump received almost eighty percent (180 out of 229) of the electoral votes from states decided by less than ten percent of the vote.


Donald Trump's narrow victories in the crucial states for 2016 allowed him to benefit from the Electoral College, but they pose the risk of a humiliating defeat this year. Because so many red states, particularly large ones, were close in 2016, a swing of a few percent of the vote to Joe Biden could dramatically reduce Trump's Electoral College tally, particularly if Biden can poach states like Ohio, Georgia, or Texas. If the most pessimistic polls for Trump are right, he could have the lowest number of electoral votes for any Republican since 1964 (currently, that is Bob Dole's loss with 159 electoral votes in 1996).

These examples show the true nature of the Electoral College. Compared to the popular vote, it punishes running up large vote totals in a few states and rewards building a coalition that can win a majority, or at least a plurality, in many states. In 2016, this meant that Donald Trump’s ability to win over the Rust Belt counted for more than Hillary Clinton’s supermajority in California, but there is no guarantee that it will not favor the Democrats in future elections, as it did as recently as 2012. Whether or not Americans support replacing the Electoral College with the national popular vote, they need to understand its true effect on our elections.

What Really Happened in the Electoral College, I

               After occurring only four times, three of them in the nineteenth century, it happened again. For decades, the possibility that the winner of the popular vote for President might lose the Electoral College was seen as a quirk of the American electoral system or a historical curiosity. Now, for the second time in less than twenty years, it was a reality.

                Sound familiar?

               In red states, Mitt Romney was an impressive vote-getter, winning by wide margins in several states that had come close to voting for President Barack Obama in 2008. He flipped Florida and Ohio, as well as three traditionally Republican states- Indiana, Virginia, and North Carolina- whose diversifying electorates gave them to Obama four years ago. Yet a string of smaller battleground states- Colorado, Iowa, New Hampshire, Nevada, and Wisconsin- as well as Pennsylvania, where the Romney campaign eschewed a serious effort until the last few weeks of the campaign, held for President Obama by margins of a few thousand votes each, giving the incumbent a 272-266 victory in the Electoral College even as Mitt Romney won the popular vote by roughly 800,000 votes.

States Narrowly Voting for Barack Obama, 2012

State

Margin of Victory as Percent of the Vote

Electoral Votes

Florida

0.88

29

Ohio

2.97

18

National Popular Vote

3.86

--

Virginia

3.87

13

Colorado

5.36

9

Pennsylvania

5.38

20

New Hampshire

5.58

4

Iowa

5.81

6

Nevada

6.68

6

Wisconsin

6.94

10

In 2012, The Electoral College Favored Obama

In the real world, President Obama won both the popular vote, by a margin of 3.86%, and the Electoral College, by a vote of 332 to 206.* However, a swing of about 2.2% of the vote, if it were proportionally distributed across the states, would have produced the scenario described above. President Obama won two states, Florida and Ohio, more narrowly than the popular vote, and a third, Virginia, by almost the same percentage as the popular vote. If those three states had switched to Mitt Romney, Obama still would have won the Electoral College, albeit by a narrower margin. To win the Electoral College, Romney needed a fourth state to change- and the next closest, Colorado, voted by a full percentage point and a half more for Obama than the national popular vote. Mitt Romney could have won the popular vote by up to 1.5 percent- over 1.9 million votes out of 129 million cast- and still lost the Electoral College.

Quantifying Trump’s Performance in the States

So, how did the Electoral College go from a slight bias toward the Democratic candidate to an even wider one for the Republican?

Donald Trump improved on Mitt Romney’s national popular vote total by about 3.3 percent, from just under 61 million to just under 63 million. Hillary Clinton received about 62,000 fewer votes, or 0.1 percent, than Barack Obama did in 2012. Minor party and write-in votes surged from 2.4 million in 2012 to 8.3 million in 2016, for an overall increase in voter turnout of about six percent nationwide.

If these effects were evenly distributed across the country- if Donald Trump had gained on Mitt Romney by 3.3% in every state, and Hillary Clinton had dropped off by 0.1% in every state- Clinton’s popular vote plurality would carry every Obama state from 2012 except Florida, enough for her to win.  The map and chart below compare the actual results to that theoretical result.

 

 


States with Overperformances Above 100,000 Votes

State

Predicted Result

Actual Result

Overperformance

Arizona

Trump +250,949

Trump +91,234

Clinton +159,715

California

Clinton +2,843,712

Clinton +4,269,978

Clinton +1,426,266

Georgia

Trump +376,564

Trump +211,141

Clinton +165,423

Illinois

Clinton +809,465

Clinton +944,714

Clinton +135,249

Indiana

Trump +317,619

Trump +524,160

Trump +206,541

Iowa

Clinton +66,529

Trump +147,314

Trump +213,843

Kentucky

Trump +445,077

Trump +574,117

Trump +129,040

Massachusetts

Clinton +691,413

Clinton +904,303

Clinton +212,890

Michigan

Clinton +375,602

Trump +10,704

Trump +386,306

Minnesota

Clinton +172,271

Clinton +44,593

Trump +135,396

Missouri

Trump +309,739

Trump +523,443

Trump +213,704

New York

Clinton +1,907,151

Clinton +1,736,585

Trump +170,566

Ohio

Clinton +73,920

Trump +446,837

Trump +520,757

Pennsylvania

Clinton +216,689

Trump +44,292

Trump +260,981

Tennessee

Trump +551,780

Trump +652,230

Trump +100,450

Texas

Trump +1,418,770

Trump +807,179

Clinton +611,591

Utah

Trump +513,959

Trump +204,555

Clinton +309,404

Virginia

Clinton +86,022

Clinton +212,030

Clinton +126,008

Washington

Clinton +419,562

Clinton +520,971

Clinton +101,409

West Virginia

Trump +193,676

Trump +300,577

Trump +106,901

Wisconsin

Clinton +164,040

Trump +22,748

Trump +186,788

 

Donald Trump’s overperformances were more widespread than Hillary Clinton’s. Trump overperformed in thirty-five states, while Clinton overperformed in fifteen and the District of Columbia. Hillary Clinton's overperformances were concentrated in California, where she overperformed by 1.4 million votes, and Texas, where she overperformed by over 600,000 votes. All these did was turn already blue California bluer and reduce Trump's margins in Texas, but not enough to carry the state or even make Trump devote resources to holding it. The only state where Trump overperformed by over 500,000 votes was Ohio, which flipped the state.

Among the other states where Hillary Clinton overperformed by between 100,000 and 500,000 votes, only Virginia was a battleground in 2012. Three of the others (Arizona, Georgia, and Utah) were states that are moving toward the Democrats, but not by enough for Clinton to carry them in 2016, the remaining three (Illinois, Massachusetts, and Washington) were safe Democratic states where she ran up the margins. Donald Trump, by contrast, had the good fortune to overperform where he most needed to: a group of states roughly bordered by the Mississippi Valley, the Great Lakes, and the Appalachians. He overperformed by between 100,000 and 500,000 votes in eleven states, and among those states, only New York was out of reach for Republicans. Five other states (Indiana, Kentucky, Missouri, Tennessee, and West Virginia) were already securely Republican. The other five included the four Rust Belt states, in addition to Ohio, to switch from Obama to Trump (Iowa, Michigan, Pennsylvania, and Wisconsin) and Minnesota, where Donald Trump came within 45,000 votes of being the first Republican to carry the state since Richard Nixon’s 1972 landslide.

Hillary Clinton’s overperformances made blue states bluer or red states closer; Donald Trump’s made blue states red.