A Proposal To Improve the Electoral College

by Alan F. Zundel

As most U.S. voters are unhappily aware, the President of the United States is elected by the electoral college, not by the popular vote. In the 229 years since the U.S. Constitution was adopted this feature has been the target of a lot of complaints. I’d like to propose for discussion a way to improve how the electoral college works.

What’s the Problem?

Here’s how the electoral college now works in almost all of the states:

  1. The voters vote for President in the November election.
  2. Whichever candidate got the most votes in a state gets to have their proposed electors be the state’s official electors.
  3. In mid-December all the states’ official electors vote as a group, or “college,” on who will be President. If a candidate gets a majority of the electoral votes (which almost always happens) that candidate is elected.

Two of the chief complaints about this system are:

The “winning loser” problem. This is when the electors elect a candidate who didn’t get the most popular votes nationwide. This happened most recently in 2000 (Bush vs. Gore) and 2016 (Trump vs. Clinton).

The “spoiler” problem. This is when a candidate outside of the two major parties is accused of “spoiling” the chances of a major party candidate to win a close race in an important state, thus affecting the ultimate outcome. Ross Perot in 1992, Ralph Nader in 2000, and Jill Stein in 2016 have all been accused of being “spoilers.”

Notice that these two problems seem to be linked: the years in which a candidate was a “winning loser” were years in which there was a purported “spoiler” candidate.

How a Less Popular Candidate Wins

The reason for the “winning loser” problem is the magnifying effect of giving all of a state’s electors to a candidate who might have only a slight edge in popular votes in the state.

A simplified example: California has about 14 million voters and 55 electors. Utah has about 1 million voters and 6 electors. Suppose the Democratic candidate wins in California with 51% of the popular vote, gaining all 55 of its electoral votes, and the Republican candidate wins in Utah with 85% of the popular vote, gaining its 6 electoral votes. Here are the results:

Democrat: 7,290,000 popular votes (7,140,000 in CA and 150,000 in UT) and 55 electoral votes.

Republican: 7,710,000 popular votes (6,860,000 in CA and 850,000 in UT) and 6 electoral votes.

The Democrat is behind in popular votes, but way ahead in electoral votes, because a slight popular vote win in a state with lots of electoral votes gets you more than a big popular vote win in a state with fewer electoral votes.

So in “swing states” (states in which the vote is close) with lots of electoral votes, a slight edge for one candidate can mean victory in the electoral college even though that candidate is behind in the popular vote nationwide.

And the presence of a “spoiler” candidate makes a big difference when the vote is close, because it may give that edge to one of the major party candidates. This is why the “winning loser” and “spoiler” problems tend to crop up in the same elections.

My Proposal

As the electoral college is set up by the U.S. Constitution, it can only be abolished by a constitutional amendment, which is a very difficult process. A way around this is to change the way states select their electors. This requires only a change in state law.

My proposal would attack the “spoiler” problem directly as a way to address the “winning loser” problem indirectly. I propose a slight change in step 2 of how the electoral college now works as outlined above, which requires only a change in state laws. The new step 2 is actually composed of two steps, so I will call these steps 2A and 2B. It would work like this:

1.(No change.) The voters vote for President in the November election.

2A. All of the candidates’ proposed electors in a state vote on which candidate’s slate of electors will become the official electors in their state. However, the weight of each proposed elector’s vote would be adjusted according to the percentage of the popular vote their candidate received in the state. Thus if a candidate received 40% of the popular vote in the state, the votes of that candidate’s proposed electors would each count for 40% of a whole vote.

2B. Whichever candidate gets the highest vote in step 2A gets to have their proposed electors be the state’s official electors.

3. (No change.) In mid-December all the states’ official electors vote as a group on who will be President. If a candidate gets a majority of the electoral votes (which should still almost always happen) that candidate is elected.

This change creates an opportunity for a “spoiler” candidate’s proposed electors to throw their vote in step 2A to one of the major party candidates rather than their own candidate. In most cases they wouldn’t want to do this, and the outcome would be the same as if step 2 had not been changed. But in cases in which delivering their vote to one of the major party candidates would make a difference in the ultimate outcome of the Presidential election, they could save their candidate from “spoiler” accusations by delivering their state’s electoral votes to the major party candidate that their voters prefer.

The new law should also require candidates to announce ahead of the election whether their proposed electors might choose to throw their vote to another candidate, and if so which one. This would make it clear that the vote switch to another candidate represents the will of those voters who voted for the candidate endorsing the switch, and thus the outcome would more closely represent the will of the majority in the state.

This change would have the biggest effect if instituted in swing states, but even in other states it could avoid potential “spoiler” problems, thus give voters greater freedom to vote for candidates outside of the two major parties if they choose to do so. That in turn creates incentives for major party candidates to appeal to voters they might otherwise take for granted, because voters would be able to turn to candidates who address particular issues if the major party candidates fail to do so.

And if the system is shown to work smoothly in other states, there would eventually be pressure on swing states to adopt it as well.

What About the National Popular Vote Idea?

The most prominent reform proposal for changing the working of the electoral college is the National Popular Vote Interstate Compact. This would have states agree to choose their electors based on whichever candidate got the most popular votes nationwide, not whichever got the most popular votes within each state. The compact would not go into effect until the states adopting it had enough electoral votes to obtain a majority in the electoral college. As of May 2018, it had been adopted by 11 states and the District of Columbia (which has 3 electoral votes), representing about 64% of the electoral votes needed.

This proposal would solve the “winning loser” problem, but not the “spoiler” problem, which would be even worse. Instead of being limited to just the states where the race was close, the “spoiler” problem would apply to the nationwide vote, with the result that any vote recount to identify miscounted votes would have to be conducted in every part of the nation, not just in particular locations. From the standpoint of practicality and delay in confirming the winner, that would be a nightmare.

It would also create a serious disincentive for voters throughout the nation to vote for non-major party candidates. This disincentive, which already exists in most of our elections, is anti-democratic because it discourages voters from voting for the candidates they truly prefer and artificially inhibits the chances for non-major parties to grow.

There is also the additional problem that under the compact a state will at some point have to award its electoral votes to a candidate who did not win the popular vote in that state. The first time this happens I would expect there to be a backlash by voters within that state against the compact, and I suspect it is a reason more states have not adopted it.


If swing states adopted my proposal, the incidence of “spoiler” problems in the Presidential election should be lessened and thus the chances of the “winning loser” problem occurring would also be lessened. Voters would feel more free to vote for the candidates they truly prefer and major party candidates would have to address issues that they otherwise might be tempted to ignore.

In other states the adoption of this proposal would have little effect on the outcome of the election, but it would prevent potential “spoiler” problems from arising in that state in the Presidential election, increase voters’ sense of freedom to vote for who they truly prefer, and create an example that swing states may want to emulate.

Your thoughts?