by George Kopf

Imagine that you owned a car that worked most of the time.  Nine times out of ten, your vehicle would back down the driveway and take you where you wanted to go.  But sometimes, on an average of one trip in ten, you’d back the car down the drive, into the street, check your mirrors, put the transmission into drive and suddenly the car would shoot backwards someplace you didn’t want to go such as your neighbor’s trash bin or mail box.  What would you do with that car?

I expect your response would be something like “I’d get another car” or “I’d get the car fixed.”  If you responded by saying “Oh, that’s the car for me.  I wouldn’t change a thing!”  you may have some problems that go deeper than a hypothetical car’s weird transmission.  The point of this  parable is that it exposes the Electoral College’s behavior and our attachment to it.

Since the U.S constitution was ratified there have been 58 presidential elections.  The Electoral College has distorted the results of nearly all of them. On four occasions — twice in the nineteenth century (Hayes in 1876 and Harrison in 1888) and twice in the twenty-first century (Bush in 2000 and Trump in 2016) — the Electoral College put the popular vote loser into the White hHouse.  That’s a plus nine percent fail rate, but the Electoral College’s overall score is actually lower because it made a total hash of two early elections (1789 and 1800, more details later.)  There were no good old days for the Electoral College.  It has been a faulty institution since its inception.  Not only does it distort our answer to one of the most important questions our country asks  us  — Who do we wish to be our president? — but it is based on ideas in which very few of us believe.

I harbored suspicions about the Electoral College for some time.  To explore them, I built a spreadsheet contrasting the Electoral College’s recent record based on popular votes to check for distortion. 

Here’s what our elections since World War II  looked like after I applied the data:

Year	Winner	Popular Win Margin	Popular Votes	Electoral Votes	Distortion
		as % of vote	as % of total popular vote	as % of total electoral votes
1940	Roosevelt	9.96%	54.70%	83.60%	28.90%
1944	Roosevelt	7.50%	53.30%	81.25%	27.95%
1948	Truman	4.48%	49.40%	57.10%	7.70%
1952	Eisenhower	10.85%	54.90%	83.20%	28.30%
1956	Eisenhower	15.40%	57.40%	86.20%	28.80%
1960	Kennedy	0.17%	49.70%	56.40%	6.70%
1964	Johnson	22.58%	61.10%	90.33%	29.23%
1968	Nixon	0.70%	43.40%	55.94%	12.54%
1972	Nixon	23.15%	60.70%	96.65%	35.95%
1976	Carter	2.06%	50.00%	55.20%	5.20%
1980	Reagan	9.74%	50.40%	90.89%	40.49%
1984	Reagan	18.21%	58.80%	97.58%	38.78%
1988	Bush	7.72%	53.40%	79.18%	25.78%
1992	Clinton	5.56%	43.00%	68.77%	25.77%
1996	Clinton	8.51%	49.20%	70.44%	21.24%
2000	Bush	-0.51%	48.35%	50.37%	-2.02%
2004	Bush	2.46%	50.70%	53.16%	2.46%
2008	Obama	7.27%	52.90%	67.84%	14.94%
2012	Obama	3.86%	50.90%	61.71%	10.81%
2016	Trump	-2.09%	48.89%	56.50%	7.61%
2020	Biden	4.46%	51.30%	56.87%	5.57%

Glancing at the data, you don’t have to be a math major to see two things.  When a presidential candidate wins by a comfortable margin, more than a few percentage points, the Electoral College rewards him with a huge win.  But in a tight race, when the margin of victory narrows to one point, the system gets very volatile and is capable of transforming a loser into a winner.

Let’s look at some specific outcomes:

In 1948 Harry Truman had to fend off challenges not only from Thomas Dewey, but also Strom Thurmond who ran on a state’s rights platform and a progressive, Henry Wallace.  This narrowed Truman’s popular victory to slightly under 50%.  The Electoral College amplified this to 57%.  That’s a big number, in the elections just before and after 1948, the Electoral College gave Roosevelt and Eisenhower much larger victory margins.  John F. Kennedy’s experience in 1960 was almost exactly the same as Truman’s in 1948; JFK barely slipped by.  In these cases the Electoral College reflected the popular vote.

The volatility that the Electoral College injects into the political system is very visible in the contrast between the 1976 and 1980 elections.  In 1976 Jimmy Carter beat Gerald Ford by a tiny margin that registers as 50% of the popular vote.  In 1980 Ronald Reagan beat Carter by similarly slim margin — 0.04 % of the popular vote– but look how differently the Electoral College reacted.  Reagan’s 4/10 of one percent improvement over Carter’s 1976 performance generated an electoral landslide that buried Jimmy Carter under 91% of the electoral votes in 1980.  Now that’s a lot of volatility!

Nevertheless, from the second world war to the close of the twentieth century, the Electoral College managed a minimally acceptable performance.  In spite of weird numbers, winners won those elections and losers lost.  The system worked o.k, until it didn’t.  

In 2000 and 2016 democracy got burned.  The Electoral College flipped the popular results — thus the negative numbers on my chart.  If you are old enough to care about such things, you have already been exposed to a lot discussion about those elections so I’m not going to say anything more about them. 

Instead, I want to ask you a question. Why do we put up with this?  Why do we allow the Electoral College to add so much volatility into our political decisions?

At this point, essays  by political scientists on this subject dive deep into the college’s constitutional origins, and particularly commentaries by Hamilton, Madison, and Jay collected in the Federalist Papers.  I encourage you to read them. But my specialty area is writing briefing memos so I will bypass extended quotations in favor of commenting on the mechanics of what is happening and how we can get our electoral car fixed.  First I need to come clean about one of my biases.  I believe in one-person, one-vote elections.  I do not think it is fair to favor or disfavor voters based on who they are or where they live.  In other words I’m a small “d” democrat who believes in inclusive popular elections.

The founding fathers, including Hamilton, Madison and Jay, were not small “d“ democrats. They had their reasons. Images of mob rule made them nervous. They sought to insulate politics from quick decisions  by popular majorities. They were for the most part  small “r” republicans and small “f” federalists who wanted the states to decide the outcome of presidential elections. The states could decide how to feed information about the popular vote into their decisions.  The states would send their result to Congress, and Congress would count the electoral votes and register the overall decision. That’s what the founders wanted; that’s how they set it up, and that’s how it works today.  I encourage you to explore the bios, backgrounds, motivations and beliefs of the founders on your own time.  They were great men who gave us a great and enduring system of government.  But now I want to look under the hood.

When you vote, your vote is recorded and counted.  Digitalization is making our elections increasingly secure and honest.  The mind-numbing numbers of TV ads and internet messages in the runup to presidential elections encourage us to support candidates running in a nationwide contest, but we vote in 50 different elections administered by each state, on a state by state basis.  In that sense, there is no national popular election. We speak and think as if there were one national election, and I built my spreadsheet on that basis. But on this point we are indulging ourselves with an illusion.  What Congress counts in deciding the winner of the presidential and vice presidential elections is a compilation of state outcomes.  And in 48 of the 50 states what the popular vote decides is who wins the state on an up or down, winner takes all of basis. Trump won Texas. All of it. Biden won Michigan.  All of it.  There are no fractional outcomes. 

So what do the candidates win?  They win a package of electoral votes equal to each state’s congressional delegation.  Each state has two senators and at least one representative in the House of Representatives.  State representation in the house is determined by the census every ten years.  The number of senators is a constant – two per state. Since 1960, there have been a total of 538 electors in the Electoral College — 100 electors representing the senatorial delegations, 435 electors representing house delegations plus 3 electors representing the District of Columbia.  Five hundred thirty eight divided by two equals 269; that’s the source of the magic number of 270 electoral votes to win that we hear so much about before presidential elections. So just to be clear, when Congress tallies the result of presidential elections, it counts the total of each state’s electoral votes for the president and vice president.  The states do not split up their results according to the popular vote.  The winner at the state level gets all of the electoral votes that a particular state has to offer.

The political result of this tidy, arithmetic system is a complete mess.  We’ve already seen how the volatility factor distorts the voters’ will and can actually flip the results. This is driven by a secondary aspect of the system that is particularly unfair and unarguably inequitable.  Because the 100 electors representing the states’ senatorial delegations is a mathematical constant, a fixed number that doesn’t change, states with small populations have a greater impact on outcomes than states with large populations. This is not a minor technicality.  It’s a very big deal.

Wyoming has a population of 578,759.  The District of Columbia’s population is 692,683.  California has 39 million folks and Texas has 29 million.  In terms of Electoral College representation you can be a gun totin’  Republican from Dallas  or a chardonnay sippin’ Democrat from San Francisco,  but your vote will be worth only 25% a vote from D.C. — a place that doesn’t even have senators but is represented in the Electoral College as if it did —  or 20% of a vote from Wyoming. Your politics won’t make any difference.  The current system supersizes voters from lightly populated states and makes pipsqueaks of voters from highly populated states. This “senatorial” aspect of the Electoral College distorts the results for all the states, north or south, red or blue.  

In sum, what we have is a national double whammy. First, the Electoral College processes popular election results through a system that turns individual votes into up-or-down, state-by-state results.  Secondarily, those decisions are distributed in favor of less populated states.  That is a long way from a one-person, one-vote system.

Earlier, I commented that the Electoral College got off to a bad start by making a hash out of two early elections.  Here’s what happened.  In the original wording of the constitution, the founders forgot to specify that the states would need to report two sets of electoral college results to Congress, one for the President and one for the Vice President.  So in 1796 John Adams became our second President and Thomas Jefferson, the runner up, became Adam’s vice president.  The problem was that Jefferson had run for president as Adam’s opponent.  They were rivals, very different people from different regions and different political parties who respected, but didn’t particularly like, one another.  Fortunately, they were  able to work together without provoking a constitutional crisis.  But can you imagine such a thing in 2016  if Donald Trump were forced to work with Hillary Clinton?  Talk about volatility!

By the way, did you know that John Adams and Thomas Jefferson both died on July 4, 1826 on the fiftieth anniversary of the adoption of the U.S. Constitution?  

Back on point, the 1796 election foreshadowed a constitutional crisis that occurred in 1800.  In that election, Jefferson won the same number of electoral votes as his vice presidential running mate, Aaron Burr.  Per the constitution, the matter went to the House where the House decided on Jefferson for the presidency, but only after a painfully protracted 36 ballots. The Twelfth Amendment (1803) patched things up just well enough for the Electoral College to go on inflicting screwy results on us for another 200 years.

Obviously, I’m not a big fan of the Twelfth Amendment, but it is notable that in 1803 the systemic response to an obvious problem was to amend the Constitution.  In our era, we too frequently describe the constitution in quasi-religious terms like “God given.”  “Originalists” , who sound like ultra-orthodox rabbis, insist we mustn’t tamper with a word of such a “sacred” document.  This is not right. The Constitution is not the Bible.  It is a wholly man-made political  document.  As such, its survival depends on its ability to change, to be amended.   

Amending the Constitution is a complicated process because it’s an important process, but it can be done.  A century ago, my grandparents and their friends were young people who idealistically thought society would be better without booze.  After a decade under the Eighteenth Amendment (1919) they realized they were wrong, that the unintended consequences of prohibition outweighed its benefits so they amended the Constitution a second time (1933) to get rid of it.  And between those two amendments, they also amended the Constitution twice again to give women the right to vote (1920) and to clear up questions about  presidential and congressional terms of office (1933) . Clearly, our grandparents  were members of  a generation in a hurry and they got things done.  They were also blessedly not overly introspective.  I don’t remember any excessive hand wringing about whether they had done the right thing.  What I do remember is that my maternal grandmother looked forward to enjoying a cocktail before dinner.

I promised that I wouldn’t quote the founding fathers at length, so in closing I will keep my promise by quoting only a fragment of a sentence by Alexander Hamilton on “The Mode of Electing a President” (Federalist Paper No. 69). Hamilton wrote that “that the true test of a good government is its aptitude and tendency to produce a good administration.”

The Electoral College does not pass Hamilton’s test.  Let’s fix it or get rid of it.  We can replace the Electoral College with popular national elections for President and Vice President.  Or, if we want to continue the tradition of Congress counting electoral votes, we can modify the Electoral College by amending the constitution to eliminate the 100 “senatorial” votes and distribute the remaining votes on the basis of the popular election results.  Personally, I don’t think the Electoral College worth another conservation effort.  But either way, in its present format, it’s got to go.