It was the scariest of times, it was the stablest of times.

Contemporary American politics offers an unsettling study in contrasts. On the one hand, Donald Trump’s lies about a stolen presidential election in 2020 and his attempts to undo the results of that contest, culminating in the violent storming of the Capitol on Jan. 6, 2021, provide blaring warning signs that our democratic system is in peril. Far from turning on Trump, a large portion of his party has continued his project: In 2021, Republican state legislatures passed new restrictions on voting access while attempting to seize control of the levers of election administration; meanwhile, GOP congressional leaders moved to isolate Republican lawmakers most critical of Trump’s conduct and claims. And the House’s Jan. 6 committee continues to unearth evidence that Republican House members schemed with the White House to overturn the election.

Worries about the state of American democracy didn’t begin when Trump rejected the election’s results — indeed, they predate his entrance into politics. For the last two decades, analysts have connected dysfunction in governance to deepening party polarization, marked by an asymmetrical Republican shift toward procedural hardball and extremism. Trump’s rise both extended and accelerated a disturbing trend. When he trafficked in authoritarian rhetoric and brazenly mixed personal and public power — while steadily consolidating the loyalty of his party — analysts portrayed it as a lesson in “How Democracies Die” and “How to Lose a Constitutional Democracy.” As nearly 200 scholars with relevant expertise warned last summer: “Our entire democracy is now at risk.”

But for all the alarm and political tumult caused by recent developments in the Republican Party, voting behavior has not changed in response; it’s shown remarkable stability and continuity with patterns established at the outset of the century. Trump himself, in 2016, probably performed worse than a more conventional GOP candidate would have — reflecting what three political scientists termed the “Trump tax” in lost votes that year — but the degree of underperformance was quite modest. In 2018, boosted by historic turnout, Democrats had a banner midterm election year, but their 40-seat gain in the House was still comparable to past “wave” elections for the party that didn’t control the White House. (It was smaller than what Republicans gained in 2010, the year of the tea party revolt, though with a bigger popular-vote margin owing to Democrats’ less-efficient distribution of voters across districts.) The 2020 election was a chance for the public to render a decisive verdict on Trump and his strongman tendencies. It featured further turbocharged turnout but stunning continuity with 2016 in the overall distribution of results.

The electoral record shows that Trump’s presidency mainly accelerated patterns of polarization by age, education, geography and religiosity. (One of the important exceptions involved a shift in Trump’s favor: his gains among Hispanic voters.) President Biden and his party’s current grim poll numbers and poor performances in November’s gubernatorial elections, meanwhile, confirm that electoral dynamics continue to follow normal patterns after Trump, with the presidential party suffering at midterm.

For those paying attention to the news, the concurrence of both realities — democratic doom and electoral business as usual — can’t help but feel dissonant, as one hammer blow after another to the ordinary course of politics leaves nary a mark on election outcomes. If citizens in a democracy don’t shift their voting patterns when the standard-bearer of one party rejects the results of an election, what would lead them to change their minds?

The simultaneity of democratic peril and electoral politics as usual can be partly explained by one of the most potent forces in modern politics: partisanship. First, an intensified sense of partisan teamsmanship allows political elites to go for broke in pursuit of their enemies’ defeat — helping to explain why Republican insiders might condone Trump’s authoritarian bent in the service of sticking it to the “Democrat Party.” And second, partisan identity powerfully anchors ordinary voters’ choices at the ballot box, even when a particular candidate transgresses one norm after another. Whatever the causes, today’s mixture of ordinary and extraordinary politics underscores the notion that the American republic is more likely to perish with a whimper than with a bang — not through violent insurrection but rather through the piecemeal subversion of another closely fought election.

Putting the fate of democracy momentarily aside, the abiding sturdiness of electoral dynamics certainly augurs gloomy tidings for Democrats this year. Two governor’s races last fall offered an early taste. The party’s candidates lost in Virginia and won in New Jersey, but they got, respectively, 12 and 13 points less than Biden’s 2020 vote share — confirming that the basic laws of political physics in normal times are still operational. (The time-honored pattern of losses for the presidential party in off-year and midterm elections can be attributed to a combination of depressed turnout for the president’s party, compared with its presidential election performance, and predictable shifts in public opinion against that party’s policy direction, shifts that are sometimes compared to the response of a thermostat to temperature changes.) There are plausible reasons to expect the electoral environment to improve for Democrats by November: Inflation might subside, as gas prices have already begun to, while the economy continues to grow, and congressional Democrats might still break through the painful negotiating stage to get a Build Back Better package passed. But there’s no reason whatsoever to think that any such reversal would be enough to save the party from significant election losses in the fall, barring extraordinary events.

So should Democrats scream from the rooftops that they are the republic’s last hope? Maybe not. Enduring electoral stability amid democratic peril suggests that messaging focused on democracy and its vulnerability has limited voter appeal. Tactically, such arguments may come across as too abstract to be motivating, particularly to swing voters, who are disproportionately less engaged with politics than partisans are. A recognition of this reality is clearly what inclined congressional Democratic leaders, in the 2018 midterm campaign, to emphasize Trump-free “kitchen table” appeals such as prescription drug pricing and to initially resist impeaching Trump — to the bewildered frustration of many progressives.

In a more profound sense, many Americans’ commitment to democracy and the principles necessary to sustain it may simply be thinner than commonly thought. Recent survey experiments have confirmed what election results suggest: that Americans just don’t feel all that compelled to punish politicians for transgressing democratic norms, especially if they have to make partisan and ideological trade-offs to do it.

For example, a survey by the organization Bright Line Watch, founded by political scientists concerned about the future of democracy in America, presented participants with hypothetical candidates for “an upcoming election.” The candidates were given randomly assigned genders, policy positions, party affiliations and stances on democratic values (such as a commitment to keeping law enforcement investigations of politicians free of partisan influence), among other characteristics. Respondents were about 19 percentage points more likely to vote for a candidate of their own party than the opposite party, regardless of their stances on policy issues and democratic values. Republicans and Democrats alike did disfavor candidates who signaled a clear willingness to politicize investigations — but only by four and five percentage points, respectively, all other things being equal. Candidates’ positions on taxing the rich, on the other hand, had twice as much influence for both parties.

And actual voters may be even less likely to punish politicians for violating norms. The academic findings stem from studies in which anti-democratic positions are stipulated as fact, whereas in the real world such information is heavily mediated through partisan communication. An October Grinnell College poll showing that a higher share of Republicans than Democrats believe that American democracy is in danger — 71 percent vs. 35 percent — suggests the influence of such mediation, as does a more recent Washington Post-ABC poll showing a significantly higher proportion of Republicans than Democrats expressing doubt that their vote will be counted in this year’s midterm elections: Large numbers of Republicans, after all, have imbibed false claims about rampant Democratic voter fraud.

In the face of genuine efforts to subvert practices intrinsic to democracy’s functioning — above all the acceptance of electoral outcomes and the transfer of power — voters’ general lack of responsiveness is not just troubling in its own right. By showing that there is little electoral penalty to be paid for such efforts, citizens give politicians an incentive to pursue these strategies and ease the path to office for anti-democratic actors. As Washington Post columnist Perry Bacon Jr. concluded after analyzing what he called the “fairly normal” results in Virginia and New Jersey: “With U.S. democracy on the precipice because of the extremism of the current GOP, everyone needs to understand that normal could well be catastrophic.”

That’s why Democrats staring down an electoral buzz saw should pursue reforms that bolster democracy while they still can, suspending or ending the filibuster in the process if necessary. Dropping the filibuster to pass the Freedom to Vote Act should be a no-brainer, for example: The bill would prohibit extreme partisan gerrymandering; provide baseline standards for voter access, including automatic voter registration and uniform early and mail voting standards; and offer (admittedly modest) protections against state-level election subversion, such as empowering election administrators to sue if they are removed from office for political reasons. When it comes to safeguarding democratic procedures, there’s no alternative but for leaders to lead.

But ultimately, of course, the principal responsibility resides with the GOP. As we’ve seen, many Republicans also perceive looming threats to democracy and to their survival as a party, in the form of voting rules that they consider overly permissive, inviting allegedly rampant fraud. The supposed need to thwart such threats serves to motivate and justify anti-democratic tactics. Recent contests suggest, though, that Republican electoral fears are as unwarranted as the evidence of fraud is scant: They confirm that the party remains perfectly competitive at the ballot box. Partly, it’s true, the GOP’s ability to keep winning elections is due to structural biases that give it a bump in the Senate, House and electoral college, but it’s not entirely due to those biases. Republicans picked up 13 House seats in 2020, and Trump came within 45,000 votes of another electoral college victory. In Virginia, Republican Glenn Youngkin won an outright majority in November’s gubernatorial election in a state that had looked solidly blue.

Neither ongoing demographic change nor the recent increases in overall voter participation, it turns out, have stopped Republicans from winning elections without anti-democratic tactics. It would be helpful if more Republicans recognized that normal politics still gives them a perfectly good shot at victory — and that they don’t need to burn the house down to win power. But the party’s recent illiberal turn has deep roots, drawing on currents of extremism and procedural ruthlessness on the American right that stretch back many decades — and the very fact that the electoral punishment for transgressing democratic norms is so slight means Republicans have no need to grapple with the trade-off if they don’t wish to. And this is how electoral politics as usual might doom democracy itself.