I make no predictions about the 2024 election. It is still far away, and all kinds of dramatic shifts could occur between now and then. But there is a clear chance that Donald Trump will win. One of several paths to that outcome leads through a recession during the next six months.

I am also reluctant to predict what Trump will do if elected; I suspect he doesn’t know himself. But we should take seriously the possibility that he would do what he has been talking about lately, including directly ordering the prosecution of political opponents, invoking the Insurrection Act, building mass camps for immigrants, purging the civil service, and even attacking Mexico.

I disagree with Hillary Clinton that these events “would be the end of our country as we know it.” On the contrary, they would mark the beginning of a new phase with highly uncertain outcomes. Much would depend on how opponents respond. Now is the time to prepare for this contingency.

Trump would have significant support, including a popular base. Certain organizations and institutions would take his side, perhaps including at least one house of Congress.

But he would also face mass resistance from segments of the population and from important organizations and institutions–notably, from some state and local governments. He would quickly encounter roadblocks, which would frustrate him and his supporters. Some of his efforts might go forward, at least temporarily, which would enrage his opponents.

The result would be intense conflict, not only in Congress and courts but also potentially on the streets. I don’t think a literal civil war is likely, if only because the US military and security services would refuse to be drawn in, and it’s extraordinarily difficult to create an army from scratch. But it is common around the world to see periods of political conflict, typically labeled “unrest,” “instability,” or “disorder.” We might expect:

• constant debates about whether various institutions should make statements about recent incidents, with repercussions for members of these institutions who disagree;
• frequent crises that are permitted by existing laws, such as government shutdowns and even a debt default;
• politically motivated pardons, amnesties, and blocked prosecutions;
• prominent dismissals and resignations;
• bans and purges of ideological minorities within institutions such as universities, corporations, and publications;
• overt refusals to follow constitutionally permissible directives (e.g., state governors might resist federal mandates);
• temporary closures of schools and colleges that are political hotbeds;
• attempts to declare martial law and states of emergency at various levels;
• arrests of questionable legality;
• illegal orders that are either accepted or refused;
• Increasingly flagrant displays of weapons;
• paramilitary and revolutionary organizations, with training programs, uniforms, insignias and the like;
• large and frequent protests, some of which may involve clashes with counter-protesters or the police;
• frequent threats of violence;
• politically motivated assaults and homicides of various kinds (not only assassinations, but also quasi-accidental deaths).

I’d expect similarities to periods like the Years of Lead in Italy or The Troubles in Northern Ireland–among many other examples. In fact, we may already have entered a period like that.

I would anticipate passionate and fraught disagreements within the potential resistance to Trump. For example:

• Should the objective be to restore and protect the constitutional system as it has been, or was that system already flawed (and responsible for the present crisis) so that it needs to be changed? If it requires change, how basic and radical must that be?
• Is the Democratic Party a worthy vehicle of resistance, or even the main opposition, or is it part of the problem? This debate will be especially fraught if it looks as if Biden would have won without third party presidential candidates in 2024.
• How broad should the coalition be? It’s easy to say “As broad as possible,” but the hard questions arise when activists must consider whether to defer causes that they consider important in order to collaborate with people who are ideologically dissimilar. For instance, imagine that it is possible to draw businesses into a pro-democracy movement, but at the cost of delaying strong action on climate. Many people would balk at that tradeoff. But what if strong federal environmental action seems impossible, anyway? Would it then be worth submerging environmental goals to expand the pro-democracy movement?
• What means are appropriate–or necessary–to combat authoritarian tendencies and street-level violence?

I think the response should be massively nonviolent, and we should eschew concrete, physical violence even in the face of institutionalized injustice. Nonviolent direct action is a powerful strategy with a strong record of success. It is particularly likely to draw broad participation and to yield a stable democracy as its outcome.*

There may be times when violence is appropriate: George Washington, Abraham Lincoln, and Franklin Roosevelt all commanded armies that fought for freedom. And sometimes it is a mistake to criticize acts of violence even if you wouldn’t endorse them. In response to the Detroit riots of 1967, Dr. Martin Luther King Jr. walked a careful line. He said that crimes committed during the riots were “deplorable” but also “derivative.” He explained, “If the soul is left in darkness, sins will be committed. The guilty one is not he who commits the sin, but he who causes the darkness.” Nevertheless, King continued to defend nonviolence because he believed that it was the most powerful option, with the greatest chance of creating a better society. That argument will be even stronger under conditions of social unrest and escalating, tit-for-tat violence. Apart from anything else, as Bayard Rustin argued, political success requires the support of a substantial majority, and violence alienates people.

Nonviolence takes skill, discipline, and values, all of which that can be taught and practiced in advance of a crisis. Now is the time for practice and training.

I am discussing a threat that comes from the extreme right. This threat is not symmetrical. However, intimidation and violence may be reciprocated, and ugly behavior may spread across the spectrum; this common pattern must be resisted.

It will be crucial to promote dialogue and listening. People will need ways to exit extremist movements and be reintegrated. And we need to hear about legitimate grievances from all quarters so that they can be addressed.

Anyone who is knowingly involved in violating civil rights should ultimately be held accountable. But tens of millions of people will vote for each major party’s nominee in the 2024 election, and voters on both sides are members of our national community. As the risk of violent conflict rises, so does the need for empathy and curiosity across partisan differences.

*See also: the case for (and against) nonviolence; Why Civil Resistance Works; tools for the #resistance; timely quotes from Bayard Rustin (1965)