Two hundred years ago, the Founding Fathers made a mistake. They decided that the president of the United States should be elected by a popular vote held among the entire country’s citizens. The results of this flawed system speak for themselves. Under the popular vote, Americans have endured two centuries of elections where the presidential candidate who receives the most votes is also the winner.

This is simply too much democracy. Time and time again, the American people have shown that they can’t be trusted to choose the leader of the free world with such a simplistic electoral method. Then again, Americans would most likely overthrow a system in which they never got to choose their leaders. The only viable alternative is one where the majority only gets to choose its leaders some of the time. In other words, we need an electoral college.

Here’s how it would work: Each state would get a certain number of electors equal to their seats in the House plus their seats in the Senate. That would give each state a minimum of three electoral votes and roughly 535 votes altogether. The District of Columbia would also somehow have three electors, bringing the total to 538 electoral votes. These electors would then gather in Washington one month before inauguration day to vote for the president and vice president. Whoever receives a simple majority of 270 votes would win. This wouldn’t be quite as radical as it sounds. Similar mechanisms exist in other thriving democracies, like the Vatican and the Holy Roman Empire.

Who would choose these electors? State legislatures could decide whether to let the people elect the electors, thereby preserving an element of the old popular-vote system. Or they could save their citizens the time and money of holding a presidential election and simply choose the electors themselves. Leaving both options on the table seems like a good idea. If no candidate receives 270 votes, either because of a 269-269 tie or because the electoral vote fractured across three or more candidates, the House of Representatives will elect the president and the Senate will elect the vice president. This creates the possibility that the elected president and vice president could be political foes, which would be productive.

Such a system would have multiple benefits for our political system. First, it would treat Americans differently according to where they live. The popular vote currently gives every citizen—Californians and Texans, Hawaiians and Mainers, Alabamans and Oregonians—an equal say in choosing the president. An electoral college would solve this problem by giving Americans in some states more influence than Americans in other states. Those benefits would largely be felt by smaller rural states, whose residents deserve to wield disproportionate levels of influence in our political system.

The numbers speak for themselves. Had the country used an electoral college in the 2016 election, an elector from California would have represented roughly 712,000 people, while an elector from Wyoming would have represented just 195,000 people. That imbalance might seem unfair to some political observers, particularly among Democrats and liberals. But they are hardly unbiased in this matter. After all, they have a vested interest in defending an electoral system that would favor their preferred outcomes, and I do not.

Creating an electoral college would also reduce the political influence of Americans who live in territories like Guam and Puerto Rico. Under the popular vote, they would enjoy an equal say in choosing their president. But because electors are divvied up among states, they would lose that status under the electoral college. Maybe this is for the best. As one National Review contributor pointed out when arguing against Puerto Rico’s statehood earlier this year, the island’s three million American citizens largely speak Spanish and live under a mismanaged government. Why should their votes count equally alongside those cast in corruption-free, fiscally sound states like Illinois and Kansas?

The electoral college would simplify presidential elections in other ways, too. It’s possible that some states would choose to allocate their electors in a proportional fashion. Modern political polarization makes it more likely that they’ll adopt a winner-take-all system instead, where candidates can capture a state’s entire slate of electors with a mere plurality of the vote. Critics somehow see this as a downside. They argue that it would disenfranchise large swaths of Americans across the country—conservatives in bright-blue California and New York, liberals in ruby-red Alabama and Utah, and those who don’t quite fit in to either camp.

On the contrary, it would only effectively disenfranchise them. Those voters could still cast ballots for president; they just wouldn’t matter most of the time. And there are benefits to shrinking the electorate as well. Candidates won’t have to waste time traveling across the country and trying to appeal to as many voters as possible. Instead, they could win just by focusing on small pockets of them. Most presidential elections would likely end up focusing on a handful of “battleground” states where either of the two major-party candidates might prevail. So instead of forcing the entire country to fully participate in the democratic process, the electoral college would relegate most of that burden to just part of it.

Under the popular vote, whoever receives the most votes will always win. The electoral college would finally end this tyranny of the majority by ensuring that candidates who come in second will occasionally become president. A study published this week by three University of Texas researchers used historical election data to measure how often the electoral college would elect someone who didn’t win the most votes. They found that in races where the candidates are separated by just two percentage points, there’s a 30 percent chance of an “inversion” where the popular vote winner and the electoral vote winner are two different people. The odds of an inversion only increase as the hypothetical margin narrows.

What’s more, the researchers also found an imbalance in who would benefit from those inversions. They estimate that Republicans would win the presidency in 65 percent of cases where they narrowly lose the popular vote. Democrats will likely fearmonger about this projected result by saying that it enshrines an illiberal structural bias in the American political system, and that it could destabilize public confidence in the democratic process. But it’s an acceptable outcome to me, a neutral observer who has absolutely no larger interest in which party wins presidential elections.

After all, one of the virtues of my proposal is that it would restrain “mob rule.” I borrowed those terms from a handful of early American leaders who opposed the popular vote when it was approved at the Constitutional Convention. Is this an anti-democratic straw man? Absolutely not. After all, the United States is a republic, not a democracy, which are two mutually exclusive things. An electoral college would help prevent the presidency from being wielded by those who would use it to impose tyrannical measures like universal background checks for firearms. Instead, it would ensure that governance occasionally falls to sober, responsible leaders who can enact moderate policies like breaking up migrant families at the border and jailing their children.

Sadly, while the case for an electoral college is clear, the path to breaking the shackles of the popular vote is not. Changing the system would require getting a constitutional amendment through both chambers of Congress, and then it would need the assent of three-fifths of the state legislatures. That shouldn’t deter reformers who want to ensure they’ll get access to political power from time to time. After all, the alternative is letting the people elect whoever they want as president forever. It’s hard to imagine anything less American than that.