This past summer, a U.S. district court declared Google a monopolist. On Wednesday, the Department of Justice filed its proposed remedy. This plan—the government’s “proposed final judgment,” or PFJ—must be approved by the judge who is overseeing the case. But it outlines changes the government believes would be sufficient to loosen Google’s grip on the search market.

Notably, the DOJ has proposed that Google sell off its Chrome web browser—which currently accounts for about two-thirds of the browser market—and stay out of that business for five years. That proposal may seem righteous and effective, and stripping Google of its browser does make the government look bold. The proposal also seems to right a cosmic wrong from more than two decades ago, when the DOJ tried (and failed) to get Microsoft to unbundle its own Internet Explorer browser during a prior antitrust enforcement. This time around, the government’s lawyers insist that wresting Chrome from Google’s mitts is necessary to prevent Google from setting a default search engine for the majority of internet surfers, and pushing those same people to other Google products too. (By the same logic, the PFJ prevents Google from paying rivals such as Apple for default-search placement.)

This is a mistake. Google’s control of Chrome has surely benefited its market position and its bottom line. But Chrome might remain a boon for Google even if it’s under outside ownership. Instead, why not force Google to strip Chrome of its Google-promoting features, and let the browser be a burden rather than a tool for market domination?

In August, I argued that declaring Google a monopoly might not matter, because the company had already won the search wars. Searching the web effectively via text typed into input boxes was Google’s first and perhaps only innovation; the competitors that arose—DuckDuckGo, Bing, and so on—offered their own takes on Googling, which became the generic term for searching the web. People returned to Google because they wanted to, not just because the company had strong-armed them.

[Read: Google already won]

Google did incentivize competitors to maintain that status quo. Mozilla’s Firefox browser offers a case study. The foundation’s most recent annual report lists $510 million in royalty revenue for 2022, some of which surely comes from Google in the form of referral fees for Google searches. The PFJ appears to prohibit these kinds of payments, and whatever revenue they generate for Mozilla. If those are off the table, browser companies may end up letting users choose their own default search service. But the results could ultimately look very much the same: People who like and are familiar with Google might just choose it again.

Google built the Chrome browser in part to steer web users to its services—Search (and the ads it serves), Gmail, Google Docs, and so forth. Search was, of course, the most important of these. (Chrome was the first major browser to integrate web-search functionality directly into the address bar, a design known as the omnibox.) But over time, other Google features have become more and more entwined with Chrome’s operation. When I opened my Chrome browser in order to write this article, it presented me with a user-profile screen, strongly encouraging me to log in to my Google account, which gives Google insights into what I do online. That facility also offers me seamless access to Google Docs and Gmail, because I am already logged in.

Given that Chrome accounts for so much of the web-browser market, a more effective way to quash Google’s bad tendencies might involve sabotaging its browser rather than selling it off. Instead of making Google divest Chrome, the DOJ could have it keep the browser running (and secure) while stripping it of all the features that make Google services ready to hand. Killing the omnibox would be the boldest of these acts, because search, which basically means Googling, would no longer be presented as the default act on the web. Likewise, removing the tight Google-account integration and associated benefits for Google’s services and data collection would frustrate the company’s monopoly more effectively than a spun-off browser ever could.

Over the past several years, courageous Americans have risked their careers and perhaps even their liberty in an effort to stop Donald Trump’s return to power. Our collective failure to avoid that result now gives Trump an opportunity to exact revenge on them. President Joe Biden, in the remaining two months of his term in office, can and must prevent this by using one of the most powerful tools available to the president: the pardon power.

The risk of retribution is very real. One hallmark of Trump’s recently completed campaign was his regular calls for vengeance against his enemies. Over the past few months, he has said, for example, that Liz Cheney was a traitor. He’s also said that she is a “war hawk.” “Let’s put her with a rifle standing there with nine barrels shooting at her,” he said. Likewise, Trump has floated the idea of executing General Mark Milley, calling him treasonous. Meanwhile, Trump has identified his political opponents and the press as “enemies of the people” and has threatened his perceived enemies with prosecution or punishment more than 100 times. There can be little doubt that Trump has an enemies list, and the people on it are in danger—most likely legal, though I shudder to think of other possibilities.

Biden has the unfettered power to issue pardons, and he should use it liberally. He should offer pardons, in addition to Cheney and Milley, to all of Trump’s most prominent opponents: Republican critics, such as Adam Kinzinger, who put country before party to tell the truth about January 6; their Democratic colleagues from the House special committee; military leaders such as Jim Mattis, H. R. McMaster, and William McRaven; witnesses to Trump’s conduct who worked for him and have since condemned him, including Miles Taylor, Olivia Troye, Alyssa Farah Griffin, Cassidy Hutchinson, and Sarah Matthews; political opponents such as Nancy Pelosi and Adam Schiff; and others who have been vocal in their negative views, such as George Conway and Bill Kristol.  

[Mark Leibovich: In praise of clarity]

The power to pardon is grounded in Article II, Section 2 of the Constitution, which gives a nearly unlimited power to the president. It says the president “shall have Power to grant Reprieves and Pardons for Offences against the United States, except in Cases of Impeachment.” That’s it. A president’s authority to pardon is pretty much without limitation as to reason, subject, scope, or timing.  

Historically, for example, Gerald Ford gave Richard Nixon a “full, free, and absolute pardon” for any offense that he “has committed or may have committed or taken part in during the period from January 20, 1969 through August 9, 1974.” If Biden were willing, he could issue a set of pardons similar in scope and form to Trump’s critics, and they would be enforced by the courts as a protection against retaliation.

There are, naturally, reasons to be skeptical of this approach. First, one might argue that pardons are unnecessary. After all, the argument would go, none of the people whom Trump might target have actually done anything wrong. They are innocent of anything except opposing Trump, and the judicial system will protect them.

This argument is almost certainly correct; the likelihood of a jury convicting Liz Cheney of a criminal offense is laughably close to zero. But a verdict of innocence does not negate the harm that can be done. In a narrow, personal sense, Cheney would be exonerated. But along the way she would no doubt suffer—the reputational harm of indictment, the financial harm of having to defend herself, and the psychic harm of having to bear the pressure of an investigation and charges.

In the criminal-justice system, prosecutors and investigators have a cynical but accurate way of describing this: “You can beat the rap, but you can’t beat the ride.” By this they mean that even the costs of ultimate victory tend to be very high. Biden owes it to Trump’s most prominent critics to save them from that burden.

More abstractly, the inevitable societal impact of politicized prosecutions will be to deter criticism. Not everyone has the strength of will to forge ahead in the face of potential criminal charges, and Trump’s threats have the implicit purpose of silencing his opposition. Preventing these prosecutions would blunt those threats. The benefit is real, but limited—a retrospective pardon cannot, after all, protect future dissent, but as a symbol it may still have significant value.

A second reason for skepticism involves whether a federal pardon is enough protection. Even a pardon cannot prevent state-based investigations. Nothing is going to stop Trump from pressuring his state-level supporters, such as Texas Attorney General Ken Paxton, to use their offices for his revenge. And they, quite surely, will be accommodating.

But finding state charges will be much more difficult, if only because most of the putative defendants may never have visited a particular state. More important, even if there is some doubt about the efficaciousness of federal pardons, that is no reason to eschew the step. Make Trump’s abuse of power more difficult in every way you can.

The third and final objection is, to my mind at least, the most substantial and meritorious—that a president pardoning his political allies is illegitimate and a transgression of American political norms.   

Although that is, formally, an accurate description of what Biden would be doing, to me any potential Biden pardons are distinct from what has come before. When Trump pardoned his own political allies, such as Steve Bannon, the move was widely (and rightly) regarded as a significant divergence from the rule of law, because it protected them from criminal prosecutions that involved genuine underlying criminality. By contrast, a Biden pardon would short-circuit bad-faith efforts by Trump to punish his opponents with frivolous claims of wrongdoing.

[Daniel Block: The Democrats’ Senate nightmare is only beginning]

Still, pardons from Biden would be another step down the unfortunate road of politicizing the rule of law. It is reasonable to argue that Democrats should forgo that step, that one cannot defend norms of behavior by breaking norms of behavior.

Perhaps that once was true, but no longer. For the past eight years, while Democrats have held their fire and acted responsibly, Trump has destroyed almost every vestige of behavioral limits on his exercises of power. It has become painfully self-evident that Democratic self-restraint is a form of unilateral disarmament that neither persuades Trump to refrain from bad behavior nor wins points among the undecided. It is time—well past time—for responsible Democrats to use every tool in their tool kit.

What cannot be debated is that Biden and Vice President Kamala Harris owe a debt not just of gratitude but of loyalty to those who are now in Trump’s investigative sights. They have a moral and ethical obligation to do what they can to protect those who have taken a great risk trying to stop Trump. If that means a further diminution of legal norms, that is unfortunate, but it is not Biden’s fault; the cause is Trump’s odious plans and those who support them.

Donald Trump’s decisive victory may proclaim an unpredictable new era for American government and society, but it also reaffirmed an enduring political truth: It is virtually impossible for the incumbent president’s party to hold the White House when Americans are discontented with that president’s performance.

Americans provided Trump with a sweeping victory after a campaign in which he had darkly promised “retribution” against a long list of enemies and offered an agenda centered on mass deportation of undocumented immigrants. Trump seems within reach of winning the popular vote, becoming the first Republican to do so since 2004. Republicans, exulting in winning at least three Senate seats as well as the White House, instantly called the magnitude of the victory “a mandate”—and Trump seems sure to treat it as a license to pursue his most aggressive ideas.

Vice President Kamala Harris and her team, recognizing the threat of widespread disillusionment with President Joe Biden, tried to transform the Democratic campaign from a retrospective referendum on the performance of the administration in which she served into a prospective choice about the agenda and style of leadership she and Trump would bring to the next four years. Ultimately, she could not overcome the widespread unhappiness over the country’s current conditions. Biden’s approval rating among voters never exceeded 43 percent in any of the major swing states, according to exit polls. At least 55 percent of voters in each of those states said that they disapproved of Biden’s performance, and Trump typically won four-fifths or more of them.

Overall, despite any expectation to the contrary, the gender gap was not especially large. Harris’s inability to amass a greater advantage among women likely reflected the fact that they were at least as dissatisfied with the economy and Biden’s performance as men were, according to exit polls. Just 44 percent of women in exit polls said they approved of Biden’s performance, and nearly seven in 10 described the economy in negative terms—a view even more emphatic than the one men expressed.

Disapproval of Biden’s record and disaffection over the economy proved a headwind that Harris could not overcome. Exit polls showed that Americans remained concerned about the possible excesses of a second Trump presidency. But in their deep frustration over current conditions, they placed less weight on those worries.

[Read: How Donald Trump won everywhere]

As Doug Sosnik, the top White House political adviser to Bill Clinton, wrote in an email yesterday: “The 2024 election marks the biggest shift to the right in our country since Ronald Reagan’s victory in 1980.” The New York Times calculated that nine in 10 U.S. counties moved at least somewhat toward Trump in this cycle. A striking sign of that change was his dramatic improvement in big urban centers with large populations of Black and/or Latino voters, including the counties encompassing Philadelphia, Detroit, and Las Vegas. But Trump also improved (compared with 2020) in communities dominated by working-class white voters, such as Macomb in Michigan, Luzerne in Pennsylvania, and Kenosha and the small cities around Green Bay in Wisconsin.

Harris maintained the Democratic hold on the prosperous, well-educated inner suburbs around major cities. But in most of them, her party’s margins declined relative to its 2020 results. She slipped just slightly in predominantly white-collar areas such as Montgomery and Delaware Counties outside Philadelphia, and Oakland outside Detroit, and failed to improve on Biden’s deficit in Waukesha, around Milwaukee. The result was that in Michigan, Pennsylvania, and Wisconsin, Harris’s margins in these big suburbs were closer to Hillary Clinton’s in 2016 than Biden’s in 2020. That wasn’t enough to withstand what I’ve called the “pincer” move of Trump’s concurrent gains in the smaller, mostly white, blue-collar places and the much more diverse urban cores.

The geographic pattern of actual vote tallies for Trump captured the magnitude of the red shift more vividly than the two major surveys that try to measure voters’ behavior for media organizations: the exit polls conducted by Edison Research and the VoteCast survey done by NORC. Neither found any increase from 2020 in the national level of support for Trump among white voters; nor did the exit polls show more than minimal improvement for him among white voters in the Rust Belt states. The exit polls recorded modest improvements for Trump among Black voters, with his gains coming entirely from men, and a big improvement among Latinos. (VoteCast found solid advances for Trump among both Black and Latino voters.) In each survey, Trump made his most dramatic gains with Latino men but scored notable improvements among Latina women as well. Young voters, in both data sets, moved notably toward Trump as well.

The exit polls showed Harris winning women (of all races) by eight percentage points and losing men by 13 points. The VoteCast study similarly showed Harris winning women by seven points and Trump winning men by 10 percentage points. At that level, Harris’s lead with women was much smaller than Biden’s in 2020, and even smaller than Clinton’s advantage in 2016.

The story on the economy was similar. Two-thirds of voters in the national exit polls described the economy as only fair or poor; roughly that many expressed negative views in each of the three former “Blue Wall” states and Arizona, with discontent rising to about seven in 10 in North Carolina and Nevada, and beyond that in Georgia. Solid majorities of those economically discontented voters backed Trump in each state. So did a big majority of the roughly 45 percent of voters who said they were worse off than four years ago.

Harris did win handsomely among those who said they were better off, but they constituted just one in four voters. She also won the narrow backing of those who said their condition was unchanged. But none of that was enough to overcome Trump’s preponderant advantage among those who thought their condition had deteriorated under Biden.

Working-class voters without a college degree—many of them living paycheck to paycheck—were especially down on the economy. More than three-fourths of white voters without a college degree nationwide described the economy in negative terms—as did seven in 10 Latino voters. (An even more telling eight in 10 Latinos did so in the Sun Belt swing state of Nevada.)

The issues that Harris and the Democrats had hoped would offset economic discontent simply did not have enough bite. Two-thirds of voters in the national exit polls said that abortion should be legal in all or most circumstances, but about three in 10 of those voters supported Trump anyway. More than a quarter of women nationwide who supported legal abortion backed Trump.

[Read: Why Biden’s team thinks Harris lost]

The muting of the abortion issue was especially dramatic in the former Blue Wall states that ultimately settled Harris’s fate. In 2022, the first election after the Supreme Court rescinded the constitutional right to abortion, Democratic gubernatorial candidates Gretchen Whitmer in Michigan and Josh Shapiro in Pennsylvania each won about four-fifths of voters who supported legal abortion, while Tony Evers in Wisconsin carried about three-fourths of them. But, in a crucial erosion of that pro-choice support, Harris won only about two-thirds of those voters in Michigan and Wisconsin and about seven in 10 in Pennsylvania. A much smaller share of voters in each state said abortion should be illegal most of the time, but Trump won about nine in 10 of those.

Harris did not entirely fail at raising alarms about Trump. In the national exit polls, 54 percent of voters agreed that Trump was “too extreme.” But about one in nine voters who viewed Trump as too extreme voted for him anyway—a striking measure of their willingness to risk an uncertain future over an unacceptable present. Likewise, in the VoteCast survey, 55 percent of voters said they were very or somewhat concerned that Trump would steer the U.S. in a more authoritarian direction; yet nearly one in six of those voters supported him.

“I think that Trump has been helped by this sense that things are careening out of control at home and abroad, and it makes people more willing to contemplate the smack of authority,” William Galston, a senior fellow at the center-left Brookings Institution, told me.

Jackie Payne, the founder and executive director of Galvanize Action, which studies moderate white women, told me that according to her research, many female voters who believed Trump would improve their economic situation simply brushed aside rhetoric and proposals from him that they found troubling. “They were choosing to believe a vision of him that was aligned with what they wanted to get out of him—a strong economy—and they were absolutely discounting anything that felt extreme as disinformation or hyperbole, even if he said he would do it,” she said.

Voters around the world have reached similar judgments this year in the aftermath of the inflation that followed the coronavirus pandemic: As a Financial Times analyst pointed out this week, incumbent parties have lost ground, or lost power altogether, in all 10 major democracies that held elections in 2024. The priority voters gave to current economic conditions in their decision making followed a long U.S. tradition too. Incumbent presidents with low public-approval ratings almost never win reelection—as Jimmy Carter in 1980, George H. W. Bush in 1992, and Trump himself in 2020 demonstrated. The similar but less discussed scenario is the difficulty facing a party seeking to hold the White House even when its unpopular president isn’t running. That applied when Harry Truman in 1952, Lyndon B. Johnson in 1968, and George W. Bush in 2008 were off the ballot; their party lost the race to replace them in each case. Biden now joins that dour procession.

But the most apt precedent for this election may be 1980. Laboring under widespread discontent, including over a raging bout of inflation, Carter tried to use his campaign to shift attention to the risks he said his right-wing rival, Ronald Reagan, represented, with some success: Doubts about Reagan did keep Carter close in the polls. But in the campaign’s final days, voters decided that continuity with Carter represented a greater risk than change with Reagan—and flocked to the challenger in crushing numbers.

[Read: How Trump neutralized his abortion problem]

Voters were willing to take an even greater leap this time. Trump made almost no accommodation for voters uneasy about him. Instead, he intensified his false accusations, inflammatory racist rhetoric, and profane personal attacks. Trump has surrounded himself with extreme figures who promise a revolution in government and society.

His senior immigration advisers have promoted plans for a militarized mass-deportation operation, complete with internment camps, and the possible removal of U.S.-citizen children of undocumented adults. His party is likely to control both chambers of Congress—and in any case, the president has broad unilateral authority to set immigration policy, as well as to impose the large tariffs Trump has pledged. The Supreme Court’s conservative majority has already rendered him virtually immune to criminal prosecution for any action he takes as president. Trump is returning to the White House unbound.

Reagan’s victory in 1980 solidified a realignment in American politics that began under his Republican predecessor, Richard Nixon. Reagan cemented working-class white voters into the conservative movement’s electoral coalition—both white southern evangelical Christians and northern industrial workers in places such as Michigan’s Macomb County—who became lastingly known as “Reagan Democrats.” Those voters remain a cornerstone GOP constituency: Even four-plus decades later, they were the two groups that supported Trump in the largest numbers on Tuesday.

Many Republicans believe that Trump now has the chance to secure an equally significant shift in the party allegiance of Black men and Latino voters of both genders, who voted for him in historic numbers this week. That opportunity surely exists. But realizing it in a lasting way will require Trump and the Republican Party to maintain the support of millions of voters of color and justify their faith in him on the economy over any concern about policies such as mass deportation and more aggressive law enforcement.

Now those communities, along with all of the other Americans disappointed in Biden over the past four years, will learn whether Trump can deliver the economic benefits he promised without plunging the country into deeper acrimony.