Blog

Professor Akhil Amar at Yale Law School is one of three law professors credited with developing the idea for the National Popular Vote interstate compact (NPV), but he recently had harsh words for the compact. Prof. Amar noted that states could “…play partisan games to inflate their vote totals and their impact on the national outcome.” One of the things he mentioned is called Demeny voting.

Demeny voting allows parents to cast votes on behalf of their minor children. The idea has been around for more than a century and a half. It was proposed by the Italian priest and philosopher Antonio Rosmini, as well as the French poet and statesman Alphonse de Lamartine, in the middle of the Nineteenth Century. They wanted each male head of a household to cast votes on behalf of his wife and children. In the modern era the idea was revived by a demographer, Paul Demeny, who saw it as a way to represent the interests of children in the political process.

The appeal of Demeny voting under NPV is obvious, especially for states that would otherwise lose influence. Consider Utah, which currently has 1.12% of all the electoral votes (6 out of 538) and whose roughly 1.5 million voters in 2024 cast 0.96% of all popular votes. Utah is the kind of state the Electoral College was created to protect, but it would lose influence under NPV (at least as far as the raw math).

So what would Demeny voting do for Utah? Looking at Utah’s turnout rates, total votes cast, the number of children under 18, and making a few reasonable assumptions, it seems likely that an additional 600,000 votes or so would be cast in Utah by parents, which would bring the state’s share of the national vote count to about 1.35%, a little more than in the last election.

There’s another factor at play, of course – Professor Amar specifically referenced partisan games – state officials looking to benefit the political party that is dominant in their state. So what would Demeny voting do to the national vote count in a heavily Republican state like Utah? A lot depends on the candidate and the race – Reagan received 74.5 percent of the vote in 1984 while Trump got only 45 percent (but still won) in 2016, and in seven out of the last twelve elections Democrats have received less than 30 percent. A reasonable estimate is probably a two-to-one margin for the Republican, which means the likely net partisan benefit for a Republican candidate would be roughly 200,000 additional votes. A more lopsided margin, such as those enjoyed by Reagan and Romney, could easily push that to 300,000 more votes in the national count. And of course a state with a larger population – there’s no reason Texas or Florida couldn’t adopt this form of voting – could easily add a net of half a million or more votes to the Republican tally.

NPV supporters have some pretty unpersuasive responses to this critique, which I’ll examine in another post. But for now, legislators and others considering the implications of NPV should recognize, as Prof. Amar has, that the compact creates new and exciting – and completely legal – ways of manipulating the presidential election process. That’s something I don’t think anybody wants.