Article II of the Constitution mandates the Electoral College for electing our President. November Presidential elections are managed by States and are indirect elections in which Presidential Electors are elected to represent their state and vote for President. With the exception of Maine and Nebraska, the Presidential candidate who wins the most votes in a state wins all the Electors for that state.

While the Electoral College has worked for over 200 years, the discrepancy of results not matching the popular vote has led 2020 Democratic candidates and others to challenge this system and advocate for basing Presidential elections on the popular vote, getting behind the National Popular Vote (NPV), in which states assign their Electors to the winner of the national popular vote. This superficially appealing idea would subvert the Electoral College, keeping its mechanics while rendering its actions moot, an end-run around the Electoral College.

NPV is fraught with contradictions and problems. The primary problem with NPV is that it would undermine Federalism and upend the power of the states to run their own elections. In NPV, a state could have no votes for a candidate, and yet if that same candidate got the most votes in other states, electors would be bound to the popular vote winner, not the state’s chosen candidate. Each state bases elections on its own voting standards, rules, and procedures. If the result of Colorado’s election of Electors depends on the popular vote in Maine and 49 other states, then Colorado no longer controls its own election process. What if one state allows non-citizens or otherwise ineligible voters to cast votes? Would votes that would be ineligible to be cast in one state but cast in a different state be used to determine an outcome for that state? Would different state-level standards pass muster under the Equal Protection Clause in a Federal court?

NPV could end up magnifying the mischief possible from voter fraud. Some contend the 1960 election was stolen from Nixon by Daley’s Chicago machine swinging Illinois for Kennedy. With NPV, a corrupt voting administrator in one city could swing not just a state but the national election. How would a candidate appeal such suspicious votes?

NPV is an invitation to recount nightmares. The 2000 Presidential election hinged on Florida’s election result, with a tiny vote margin of only a few hundred; a recount of only a few counties became a month-long untenable mess. A national recount, in which every choice of Elector hinges on the outcome of that recount, would multiply that challenge by orders of magnitude. Imagine recounting 140 million votes, hundreds of counties, all at once; it could end up taking months in the case of challenges.

In making the Electoral College moot, NPV also eliminates the U.S. House as arbiter in cases where the Electors are too divided to choose. If the leading candidate gets a plurality, but under 40% (which has happened before), would every Elector be required to back that “plurality but not majority” candidate?

NPV is untenable and impractical because of problems with fraud, recounts, inconsistent election standards, and non-majority elections. But the fatal flaw of NPV is that it undermines Federalism. The Electoral College is truly Federal, breaking down our Presidential election into smaller state-level elections for Electors. NPV erases that: The smaller states would get sidelined; urban centers would dominate electioneering; individual votes would get lost in a sea of 150 million votes across the nation, and bear no relation to the state’s election results. States would no longer matter. Losing Federalism is too high a price for “fixing” Presidential elections in those occasional cases where close elections go to a candidate with slightly fewer (but better distributed) votes across our nation.

While a switch to a National Popular Vote is a flawed idea for electing our President, there are valid concerns that the Electoral College falls short in representing voters’ preferences. The main culprit is the winner-take-all method, in which a state allocates all its electoral votes to the winner of that state.

With winner-take-all, strongly partisan Red and Blue states are barely contested and have pre-determined election outcomes. Candidates focus mainly on “Battleground States,” resulting in an imbalance in turnout, attention, energy, and voter engagement.

The winner-take-all Electoral College can also lead to imbalances between an Electoral College outcome and the popular vote. Whether a candidate wins by one point or 50 points, they win the same number of Electors. Thus, a candidate who wins many states by a narrow margin can beat a candidate with large margins in fewer states. This explains Trump’s 2016 victory; in states where the margin of victory was under 1%, Trump won four critical states with 46 electoral votes, while Clinton only won one state with four electoral votes.

We can do better than winner-take-all. Currently, Maine and Nebraska are the only states to employ non-winner-take-all systems. They use the Congressional District Method, with two Electors allocated to the state-wide winner, and the remaining Electors allocated to the winner in each Congressional district.

While this method reflects more granular voter preferences, it also would be subject to biases of district gerrymandering. How can we fairly allocate Electors in a manner that reflects voter preferences without such manipulation? We propose a Mixed Proportional Voting system for a state to allocate the Presidential Electors proportionate to the percentages of popular votes statewide for each candidate:

1. Each state has two Electors above and beyond the number of Congressional Districts they have; they are awarded to the candidate that receives the most votes in the state.
2. The remaining Electors, equal to the number of Congressional Districts in that state, would be awarded proportionally to the percentage of popular votes each candidate receives state-wide.
3. To award a whole number of Electors based on statewide vote percentages, we would round down candidate popular vote totals. Candidates may have remainder popular votes, the “left over” votes after candidate vote percentages are allocated to Electors. Any unallocated Elector or Electors are awarded to the candidate(s) with the most of those remainder popular votes.
4. A candidate needs to receive at least 10% of the vote in a state to be allocated any of that state’s electoral votes.

Mixed proportional voting is a more representative way to make the Electoral College work, and would solve many problems our Presidential elections face today:

1. Every vote would become more equally meaningful, as votes in every state can move Elector votes. For example, in Texas, a candidate could win one more Elector by shifting about 300,000 votes. This would change the “Red State/Blue State” mindset to one where minority voters within each state have meaningful influence and opportunity to shift Electoral votes.
2. It would reduce “Battleground” state focus and equalize the importance of each state in the election cycle. The election focus would move to a wider battle for vote-share in every state and region as non-battleground states like South Carolina, California, New Jersey, Tennessee, etc. would have electoral votes in play.
3. It would make fraud less impactful than winner-take-all or NPV because a small number of votes would no longer tip a large number of Electors.
4. It will open the field up for third party candidates, who could win one or more Electors by crossing the 10% threshold. This might also divide electors more when there are more than two major candidates, more accurately reflecting a divided electorate.

Most importantly, adopting Mixed Proportional Voting would preserve our Electoral College and its true intent and our Federal scheme of electing the President, while embracing the election goal to make every vote count.

If it is such a great idea, why hasn’t it been adopted already? Each state’s voting majority has an incentive to maximize their influence; partisan states’ majority don’t want to dilute the winner-take-all result that benefits the majority; Battleground States have no incentive to undermine their status and influence. It’s difficult to convince each state’s majority to adopt a voting model that dilutes their power unless others do the same. It’s a voting influence prisoner’s dilemma. To address that concern, we could adopt the NPV’s compact model: Every state would propose this as a future model, to be adopted when sufficient states, i.e. a large majority of states, agree to adopt it.

Mixed Proportionate Voting is a superior way to elect our President. It is fraud-resistant, less prone to recount angst, and broadens election participation to all states and regions, including both urban and rural areas. It preserves and improves the Electoral College in both intent and meaningful operation. NPV is a threat to Constitutional Federalism, but we need to do more than say “no” to that bad idea. Let’s mend, not end, the Electoral College and move towards Mixed Proportionate Voting across the states for our Presidential elections.

Patrick McGuinness works as a technology professional and lives with his wife and children in Austin, Texas. He obtained a PhD in Computer Science from the University of Illinois in 1991.