Stronger evidence: AgainstVerdict
The evidence opposing retention of the Electoral College is stronger, more direct, and more consistent. Several studies document concrete, measurable problems with the current system. Richie and Levien (2013) show that in 2012 only 10 states received 99.6% of campaign spending, leaving 41 'spectator states' largely ignored. The 2025 'Determinative Popular Vote' study finds that outcomes can hinge on as little as 0.15% of voters nationwide in swing states, which strains the one-person-one-vote principle. The PNAS study (2020) and the Annals of the AAG study (2009) find systematic partisan bias and unequal voter power across states. The American Economic Journal study (2022) documents how often popular-vote winners have lost and the risk that this recurs. These findings come from peer-reviewed work and agree with one another.
The pro-retention evidence is thinner and less direct. The strongest item is Egorov and Sonin's NBER paper, which models the Electoral College as a fraud deterrent because swing states have divided political control. But it is a theoretical model rather than an empirical finding, and it is a working paper. The Election Law Journal (2008) and Social Sciences (2019) pieces point to legal clarity and 'bizarre scenarios' in alternatives, but as described they identify edge-case risks rather than showing that the current system performs better in practice. Some items listed in support do not clearly back retention at all. The American Politics Research study (2024) finds that support for the system tracks partisan self-interest after inversions, which weakens principled defenses of the status quo. The turnout study (2010) shows that turnout affects outcomes at several levels, which is not an argument for keeping the Electoral College.
A fair reading grants that a national popular vote could create its own vulnerabilities, such as fraud incentives in one-party states and administrative uncertainty in close national contests. These concerns are plausible but speculative. The documented harms of the current system are observed and quantified: unequal campaign attention, unequal voter power, and inversions. On the evidence presented, the case for replacing the Electoral College is better supported.
Ranked by citation count.
The debate
Elena: Let's start simple. The Electoral College has held for over two centuries. It forces candidates to build broad coalitions, and it gives us clear, legally settled outcomes. Why throw that out?
Sophia: Because "it's old" isn't an argument. Look at what it actually does. Richie and Levien found that in 2012, ten states got 99.6% of campaign spending. Forty-one states were spectators.
Elena: Campaigns go where the votes are persuadable. That's just strategy.
Sophia: It's strategy that decides whose voice counts. If you live in a spectator state, your president is picked without you. That's not a coalition-builder, that's a filter.
Elena: Fine, but a national popular vote has its own dangers. Egorov and Sonin's NBER work argues the Electoral College deters fraud, because swing states have divided political control.
Sophia: That's a theoretical model, and a working paper. It's not an observed result. Meanwhile, the 2025 Determinative Popular Vote study finds an entire election can hinge on about 0.15% of voters in swing states. How does that square with one person, one vote?
Elena: Close elections are close. Someone always decides them.
Sophia: Sure, but under the current system the deciding voters are concentrated in a few states, so everyone else's vote is functionally decoration. And it's not random. The PNAS study from 2020 and the Annals of the AAG study from 2009 both find systematic partisan bias and unequal voter power across states.
Elena: Then let's talk about legal clarity. The Election Law Journal points out that alternatives create bizarre scenarios, like national recounts.
Sophia: Edge-case risks. Plausible, I'll give you that. But "imagine a messy recount" doesn't show the current system performs better. We've already lived through messy outcomes. The American Economic Journal study from 2022 documents how often popular-vote winners have lost, and the risk that it happens again.
Elena: Inversions are rare.
Sophia: Rare isn't never, and they've happened twice in just the last quarter century. And here's the tell. The American Politics Research study from 2024 shows support for the system tracks partisan self-interest after inversions. People defend it when their side wins it.
Elena: Or people just respond to results. Turnout matters too, and that 2010 turnout study shows it shapes outcomes at several levels.
Sophia: It does, but that's not a defense of the Electoral College. Turnout matters everywhere. It just matters more equally under a national popular vote, where a vote in Wyoming or California counts the same as one in Ohio.
Elena: You're dismissing fraud incentives in one-party states. If a state can pad its totals, a national vote rewards that.
Sophia: That's a fair worry, and I'll grant it. So is administrative uncertainty in a tight national race. But those are speculative. The harms of the current system are measured: unequal attention, unequal voter power, and inversions. I'll take documented problems over hypothetical ones.
Elena: So you'd trade a system that works for one we haven't tested.
Sophia: I'd trade a system with peer-reviewed, consistent evidence of unequal treatment for one that treats every voter the same. When the studies all point the same direction, that's not a gamble. That's the evidence talking.
Elena: Hard to argue with a stack of studies.
Sophia: Then we agree. Count every vote equally.