On Friday, bestselling author and Hahvahd professor Daniel Ziblatt came to Columbia to discuss his recent book How Democracies Die. Co-authored with Steven Levitsky, it asks whether American democracy is in danger (yes) and what we can do to save it. The New York Times calls it “an essential guide to what can happen in the United States.” Staff Writer Andrew Wang calls it a guide to what’s happening right now.

I know a Hahvahd political scientist when I see one.

The older and richer a democracy is, the less likely it is to crumble. At least, that’s what history tells us. But if that’s true, then America—who did it before it was cool—should be thriving. That argument is looking harder and harder to make.

“Democracies don’t die like they used to,” says Ziblatt. In the past, it was men with guns. Ziblatt estimates that during the Cold War, three quarters of democracies fell at the hands of military regimes. But the contemporary topology of death is much more subtle: it’s Presidents and Prime Ministers who undermine constitutions by way of referendum, elections, and parliamentary legislation. And that’s hard to notice. Take Venezuela for example. In 2011, Venezuela was already 11 years under Chávez, but the majority of the country believed they were living in a democracy.

What keeps democracies stable, Ziblatt says, is its ability to keep extremists away. Parties do this through a rigorous candidate selection process where authoritarians are kept far from office. In the U.S., this gatekeeping has been effective at times; in the 1930s it stopped Huey Long and Charles Coughlin, and later, Joseph McCarthy, the antisemite Henry Ford, and the segregationist George Wallace. These authoritarians of old were not personae non gratae; Gallup poll data from the 1930s revealed 35-40% approval ratings.

This gatekeeping came at a cost. Back then, party leaders would light cigarettes and gather in “smoke filled backrooms” to strike deals that eliminated would-be-demagogues. Some inventions, such as binding primaries, gave voters more choice and party officials less. Other ones, such as the superdelegate system introduced in the 1980s, gave party leaders some of their gatekeeping power back. By the way, the Republican party doesn’t have superdelegates.

While it’s tempting to immediately declare power to the people, Ziblatt suggests that exclusion can be effective. Without it, would-be-demagogues who begin careers as political outsiders have a unique advantage. Establishment politicians, seeing their charisma, strike a Faustian bargain: they align with the demagogue to tap popular support. Then, as they say, they’ll control him. Only that hasn’t worked. Mussolini came to power this way in the 1920s. And in Weimar Germany, the nationalist Alfred Hugenberg tapped Hitler, issuing joint proclamations for their parties. When Hitler became Chancellor of Germany in 1933, his Vice-Chancellor Karl von Papen declared, “don’t worry, within two months we’ll have pushed Hitler so far into the corner he’ll squeal.”

To empower American democracy, prevention helps. But prevention becomes useless once a demagogue is already in power. Ziblatt turns to containment. Andrew Jackson, FDR, and Nixon were treated this way: by way of constitutional opposition. But as Ziblatt warns, constitutions are not self-enacting. For example, the Argentine constitution is basically a ripoff of America’s. Yet, they’ve had six military coups in the 20th century alone. “Constitutions work best when they are reinforced by unwritten laws,” Ziblatt says. In other words, norms.

Ziblatt turns to two norms that just might save democracy yet. The first is respect for the legitimacy of political rivals, as opposed to enemies. In France last year, when the Republican politician Francois Fillon failed to make it to the second round, he refused to endorse the right wing nationalist Marine Le Pen. Instead, Fillon turned the other way: toward Macron. In doing so, he moved his voting base toward the centrist politician. All this despite his fall from grace following a financial scandal. Yet, John McCain, praised across the aisle in the final years of his life for opposing President Trump, never openly endorsed Clinton.

The second norm is institutional forbearance, which means restraint from enacting one’s legal right at every possible turn. The truth is that the law lets you do a lot. The POTUS can pack the Supreme Court, pardon anyone, and rule by executive order. Congress can impeach the president on high crimes and misdemeanor charges. Senators can get paid for filibusters; from 1930-1970 there were 30 filibusters, and there are that many in each session today. By way of the constitution, Ziblatt notes that constitutional watchdogs like checks and balances can quickly become lap dogs.

Forbearance, or what Ziblatt calls “democracy’s soft guardrail,” protects us from this. Simply put, just because democracy can do something doesn’t mean it has to do it. According to Ziblatt, the beginning of institutional forbearance came at the end of the 19th century when Reconstruction ended. Parties began connecting, making deals to exercise restraint. Yet, there is a darker side to this inflection in American democracy. Ziblatt says that forbearance came at a terrible cost: the then-Republican party decided to end their commitment to racial equality to appease Southern Democrats. Only then could they begin bargaining. One could argue, then, that at the center of our democracy is an irreconcilable contradiction.

In short, institutional forbearance is no panacea to democracy. But it doesn’t follow that we should discard it, and that’s what we’re doing right now. At the end of the Obama presidency, the Republican-controlled U.S. senate refused to hold hearings for Judge Merrick Garland. This move had been unprecedented since the 1860s. As for seeing political opponents as rivals, there’s also much to be desired. Ziblatt asked us to consider Newt Gingrich’s “anti-flag” and “traitor” discourses to describe Democrats in the early 1990s. Or the Tea Party movement, or the right-wing birtherism against our first black president. To be clear, this is not just a jab at the right, but a story of how recent political trends have eroded norms. Indeed, Constitutional hardball can go both ways—consider Obama’s liberal use of executive orders—and while force can secure gains in the short term, Ziblatt argues that more consistent victories can be won without it.

While America’s softening grip on norms has made demagoguery and partisan divisions greater, the question remains: is today’s threat to democracy uncharted terrain? According to Ziblatt, yes. For one, the liberal-conservative paradigm is no longer what it once was. Politics isn’t about fighting at the dinner table over healthcare and taxes. It’s more: race, gender, and culture itself. In the last fifty years, America’s demographics have changed drastically, and in the next fifty, they’ll continue changing. The Civil Rights Movement that (arguably) secured rights for African-Americans drove them towards the Democratic Party, and southern Democrats the other way. And the overwhelmingly white and Christian ideology that dominated American life for centuries is no longer the majority. In short, democracy is under threat because the world it governs is changing. If we can rethink what a healthy democracy looks like, then it might just be saved.