If you wanted to make a surefire bet on the Super Bowl, there was an obvious one: President Trump would complain about Bad Bunny’s halftime show. And so he did.

“The Super Bowl Halftime Show is absolutely terrible” etc. etc., his social-media post began. And: “Nobody understands a word this guy is saying, and the dancing is disgusting” etc. etc. and “This ‘Show’ is just a ‘slap in the face’ to our Country” etc. etc. etc. etc. etc. etc. You could have had a pretty easy parlay on the inclusion of certain complaints.

In general, this moaning isn’t worth paying attention to. But there is one element of the complaint that’s worth calling out for its ridiculousness: the one about how “nobody” understood what Bad Bunny was saying.

He was speaking Spanish. And while perhaps no one at the White House or Mar-a-Lago speaks Spanish, lots of people in America do.

According to the Census Bureau, about 78 percent of U.S. residents only speak English at home. But about 1 in 8 speak Spanish.

Despite how central race is to Trump’s politics, it’s useful to remember that many of those who speak Spanish in the U.S. live in states that Trump won in 2024. In fact, more of them do.

In fact, four of the six states with the highest percentages of Spanish speakers voted for Trump in the most recent election.

Notice that I didn’t include U.S. territories. Puerto Rico, where Bad Bunny was from, is home to about 3 million more Spanish speakers — Americans, whether Trump likes it or not.

Plus: Millions more people watch the Super Bowl around the world. Two years ago, more than 24 million people in Mexico tuned in. I’m guessing most of them were able to figure out what Bad Bunny was saying.

It’s admittedly silly to rebut this point, in part because I’ve invested about 15 minutes thinking about something that Trump considered for less than 15 seconds. It’s just culture-war stuff, an effort to stoke his base’s sense that something is changing in a way that makes them feel uncomfortable. In a way that makes him feel uncomfortable.

The results of 2024 notwithstanding, it’s a safe bet that politicians who appreciate Bad Bunny’s performance are better attuned to the present (and future) than are politicians who portray a Spanish speaker as bizarre or threatening.

There is an uncountable number of repositories of power all across the United States. Some are huge and obvious: the federal government, major corporations, police departments, billionaires. But there are millions of small ones, too, including small businesses, community groups, churches — and the people themselves.

In theory and historically, the federal government is the aggregation of all the small individual centers of power. In recent decades, though, it’s been fair to wonder whether the federal government wasn’t instead an aggregation of the huge ones.

No matter. There is a new aggregation of the small centers of power — one manifested every single day in opposition to the federal government that was created to represent them.

We are where we are because Donald Trump saw the power and wealth of the federal government and wanted it for himself. There wasn’t obviously a path for him to get there but he stumbled forward anyway, discovering that repeating the stuff he heard on Fox News about The Elites and The Immigrants got a lot of people riled up. It also took him a while to figure out how — and, moreso, if — he could start siphoning off the country’s power, but he did.

It’s always useful to remember that Trump came to the White House not via the route of politics but via the route of capitalism. It is particularly useful to remember that in this discussion, because the voraciousness is the same. His focus has always been on accumulating as much power and influence as he could, and that’s how he approaches the presidency, too. When he spends the country’s wealth and influence, it’s on what he wants, not what benefits the country.

What’s been weird to learn is that other leaders of institutions act the same way. Some of them hoard and protect their influence like it’s money because they, too, are rich. You don’t get rich by giving your money away and (they seem to believe) you don’t get powerful by exercising your power. But, in fact, that is often how power works: using power can help build power. At the very least, power holds no value unless it is exercised, which alone is a valid reason to exercise it. If you are powerful but never use your power, you are not powerful.

I’m talking here primarily about the huge, obvious loci of power — the centers of power that exist in part to serve as checks on other centers of power. Congress vis-a-vis the presidency. The media relative to the government. Higher education in contrast to misinformation. But they often aren’t serving as checks on what’s unfolding, or are obviously and bizarrely using their power only sparingly, garden hoses aimed at protecting their own homes from the wildfire.

So it’s left to individuals, who have stepped up where institutions have not.

There’s been a lot of ruminating about why this is the case, why individual, average people have been bold but newspapers and senators haven’t been. To some extent, it’s a numbers game. There are 340 million people in the United States. About 260 million of them are over the age of 18, meaning that about 146 million adults just in the U.S. disapprove of the job Donald Trump is doing as president. That’s a big pool from which to find a few thousand who are willing to fight for democracy.

I mean, there are about 350,000 adults in Minneapolis alone. Even just in theory, it wouldn’t be hard to find a dozen who were substantially committed to opposing Trump’s deployment of power that they would make it the center of their daily activity. But in practice, thousands have. Even after two people were killed by federal agents while doing so.

What’s important about this movement is that the only power that most of these protestors have is their own. They are expending all of it — and in doing so helping to build a bigger, aggregated power. The stewards of large accumulations of power are wary and timid. Enough stewards of personal, individual power are bold and engaged that a significant counterweight to the administration has emerged.

It’s telling that so many defenders of the administration can’t accept this for what it is. The constant refrain that opponents of the administration are paid to express their opposition is generally nothing more than a bad-faith effort to delegitimize it. The idea that tens of thousands are being slipped cash to whistle at ICE officers without any evidence of such payments emerging is as dumb an argument as saying that 2020 was affected by rampant fraud — which is a reminder that the stupidity of an argument is no barrier to its adoption. Even the idea that liberal donors giving to liberal groups somehow constitutes “funding” the protests is a deflection from the fact that bespoke responses to ICE have been created on the fly, with thousands of volunteers chipping in.

Not all of the dismissiveness aimed at anti-administration protests is bad faith. In a polarized country, many Americans literally don’t know anyone who disagrees with them; to such people, the idea of sincere opposition might legitimately seem invented, particularly if they are immersed in a media environment that claims precisely that.

Part of it, too, is that the country’s ideological divide manifests in diverging views of how society works. The right focuses on individualism; the left on community. Protests are more common on the left in part because there’s an acceptance of communal action that isn’t prevalent on the right. To them, what’s unfolding in Minneapolis can seem literally inexplicable. (The overlap between this divergence and population density is probably causal; one reason kids who go to college shift left is because going to college is inherently a communal act, even when it isn’t in big cities.)

Individualism and selfishness are intertwined. And hoarding power and money is a selfish act. This is probably overly neat, but it remains true that this tendency toward hoarding is a significant reason that a far-right president has been able to erode American democracy so quickly.

But democracy isn’t solely about elections or about the exercising of aggregated national power. As the old protest chant has it, activism is also what democracy looks like. Citizens piecing together ad hoc institutions to confront oppression is democracy. People stepping up to protect the country when Congress and the Supreme Court won’t is democracy.

Democracy isn’t simply being defended. It’s being reorganized — out of unfortunate necessity, but effectively. So much so that the huge centers of power are finally starting to notice.

Photo: Protesters in Minneapolis. (Fibonacci Blue/Flickr)

ProPublica has identified the Border Patrol agents who shot and killed Alex Pretti as Jesus Ochoa and Raymundo Gutierrez, both of South Texas. There have been numerous reports that the federal agents who’ve blanketed Minnesota are not from that state or even that region; agents have been spotted wearing indicia that they are from Texas or other states.

I was curious the extent to which this federal crackdown in Democratic-voting regions of the country was being effected by officers from Republican-voting ones — how much of it was a South-to-North deployment that echoed American history more than might have been intended.

The government aggregates data on its employees, workers who, after all, are paid with public tax dollars. The Office of Personnel Management releases monthly data tables identifying information on age, education, salary range and duty stations for employees of the U.S. government. From the most recent data (covering November) we can see that Department of Interior employees are, as you might expect, spread out across the country.

Some Interior employees don’t have their duty stations recorded. But only a few — 4 percent of the total. Every other employee’s duty station is identified, the most in California.

When you do the same analysis for the Department of Homeland Security, though, the pattern is different.

Fully three-quarters of DHS employees have their duty stations redacted.

When we look at specific departments, like ICE, we see an even more dramatic shift.

The public is informed that ICE has about 2,000 employees in D.C. Everyone else? Shielded in the OPM data.

Related: ICE’s excuse for wearing masks has never actually manifested

Same with Customs and Border Protection. Of the department’s 67,000 employees, we’re informed about the duty station — D.C. — of only 5 percent.

Together, the redacted duty stations of ICE and CBP staffers constitute more than half of the DHS redactions.

To put an exceptionally fine point on it: My point is not that the duty stations of every individual must necessarily be made public, much less that their home addresses be available. It is striking, though, that we have so little insight into who is patrolling the streets of cities targeted by the administration for overtly political reasons.

---

Update: As Lauren Harper of the Freedom of the Press Foundation noted on Bluesky, this is a function of both ICE and CBP being extended new protections in 2020, during Trump’s first term in office. A memo released internally after the designation was applied to ICE explained that the designation as a “security/sensitive” agency would “ensure that OPM withholds all relevant personally identifiable information (PII) of all ICE personnel” from public information.

Photo: Items in the Border Patrol Museum in El Paso. (Jonathan McIntosh/Flickr)

The first time Donald Trump ever disparaged journalist Don Lemon on social media was during the 2016 campaign. At a rally, Trump, then the Republican nominee for president, had warned his audience that the election of his Democratic opponent Hillary Clinton would mean Supreme Court justices who would scale back gun rights.

“If she gets to pick her judges, nothing you can do folks,” Trump said. “Although the Second Amendment people, maybe there is. I don’t know.”

Lemon, then the host of a show on CNN, asked his guests how this could be interpreted in any way other than as a threat against Clinton. One guest rejected the idea, insisting that the language may have been “imprudent” but that Trump was clearly just telling pro-gun voters to turn out on Election Day. (That guest, incidentally, was Dan Bongino, the right-wing podcaster-turned-Trump FBI Deputy Director-turned right-wing podcaster.)

Trump saw a clip of the exchange and praised Bongino. Lemon, he said, “is a lightweight – dumb as a rock.”

This became a refrain: Lemon, whose coverage of Trump was often critical, was described by the president over and over again as unintelligent and biased. The CNN host — even more than other CNN hosts! — had joined an ignominious (or, depending on your politics, illustrious) group: Enemies of Trump.

On Friday morning, an attorney for Lemon reported that the journalist, now working independently, had been arrested in California. Attorney General Pam Bondi confirmed this news on social media.

“At my direction, early this morning federal agents arrested Don Lemon, Trahern Jeen Crews, Georgia Fort, and Jamael Lydell Lundy,” she wrote, “in connection with the coordinated attack on Cities Church in St. Paul, Minnesota.”

This development was striking in part because the administration had already tried to criminally charge Lemon, without success. Earlier this month, Magistrate Judge Doug Micko declined to sign an arrest warrant targeting Lemon and others. An effort to appeal that decision was also rejected.

It is also striking because of what actually occurred at the church in St. Paul. A pastor at the church is also an official with Immigration and Customs Enforcement, prompting a group of protesters to interrupt a service he was leading. Coverage of the event doesn’t suggest that it was an “attack” in the commonly understood use of the term. But that coverage itself is at the heart of the issue for Lemon: He was there, interviewing participants and the pastor, a presence that would seem clearly to be covered as a constitutionally protected activity. (Georgia Fort, another one of those identified by Bondi, was also there in her capacity as an independent journalist.)

It is tempting to view the targeting of Lemon solely through this lens, as an extension of Trump’s longstanding hostility toward the media — or, more specifically, toward any media outlet or personality that isn’t falling over itself to praise him. Trump has repeatedly used his personal and presidential power to attack reporters and media outlets. The arrest of Lemon is simply another tick on the administration’s anti-First Amendment checklist: religion, speech, assembly and the press.

Yet it is probably wiser to view the Lemon arrest as something broader. It’s not that he’s a journalist. It’s that he’s been identified as an Enemy of Trump, and the president and his base have long been howling for those perceived enemies to be punished.

Again, there’s no public evidence that Lemon was involved in the church protest in any capacity other than as a reporter. But his presence there quickly drew the attention of the pro-Trump right, centering him as a representative not only of the protest but of the broader anti-Trump movement. On social media, Lemon became an avatar for all of the disparate elements of opposition to Trump that have emerged as federal agents have swarmed Minneapolis. Countless demands that he be punished exist within that context, not within any obvious legal one.

You’re telling me they can arrest a president, but not Don Lemon? Ok.

— BLAIRE WHITE (@BlaireWhite) January 22, 2026

Arrest Don Lemon

— Adam Carolla (@adamcarolla) January 19, 2026

Whose arrests did Bondi announce? Two Black journalists and two Black Democrats who’ve sought elected office. In other words, four people who represent an assortment of things that Trump’s base views with some hostility.

To be very clear, it is not better that Trump’s Department of Justice is going after his perceived enemies rather than against journalists for doing journalism. It is far worse, because it reinforces the (already obvious) willingness of the administration to use federal power to impose penalties on critics and opponents. Not necessarily criminal penalties, like jail time, but even simply forcing people into the criminal justice system, spending time and money on attorneys, and so on.

The Trump administration will scoop up a guy in Maryland and send him to El Salvador and then backfill alleged criminal behavior in an effort to get him deported. It will shoot a mother to death in her car and then declare that she was a murderous terrorist.

It’s important that it is filing criminal charges against a reporter who was in the act of reporting. But it’s more important that it is doing so centrally because that reporter was a perceived Enemy. How else to explain the White House celebrating the arrest on social media, seeking accolades and approval from the base?

When life gives you lemons… ⛓️ pic.twitter.com/wxry0fudOj

— The White House (@WhiteHouse) January 30, 2026

The Bill of Rights included protections for journalists because the founding fathers understood that people in positions of power would find the press and its accountability systems to be annoying or threatening. What they probably didn’t foresee was a president who wanted to attack a journalist centrally because he wanted to show his supporters how eagerly he was going after their collective perceived opponents.

Photo: Don Lemon smelling cookies my wife made for him and his team back in his CNN days.

From the earliest months of Donald Trump’s second term in office, federal immigration officers have conducted their raids and arrests while concealing their identities, a notable deviation from the way in which law enforcement has long been conducted in the United States. The public grants police and other law enforcement the right to carry weapons and detain suspects; in return, they are expected to be accountable to the public that employs them.

What the Department of Homeland Security has long argued is that this is dangerous for the officers. Assaults against Immigration and Customs Enforcement and Border Patrol agents surged in 2025, we have been told. Agents must protect their identities in order to keep themselves safe from this uptick in hostility and violence.

A review of DHS and ICE press releases since January 2025, though, indicates that this theoretical scenario has never actually occurred. At no point in time has an officer been seen conducting his work, identified and subsequently attacked. While there have been threats issued against agents and incidents of off-duty harassment, there are no known incidents in which an officer was assaulted while off-duty because he was identified as a federal agent.

There appears to have been only one incident in which an officer was attacked while off-duty. That occurred in July, when an officer was shot during a robbery attempt in New York City. One of the attackers was an immigrant living in the country despite a deportation order.

In September, a sniper fired on an ICE facility in Texas, killing two detainees and injuring another. No officers were struck. A clip found at scene had “ANTI-ICE” written on it; the alleged shooter appears to have had mental health issues.

That this attack occurred at an ICE facility is important. The other assaults documented in DHS and ICE press releases all occurred when agents were either working at an ICE facility or attempting to execute an operation. In other words, they occurred at times when they were obviously identifiable as ICE agents. There was no tracking them down to ID them; the incidents occurred while they were at work. Many of the documented assaults involved immigrants who were hoping to evade detention. Had the agents not been wearing masks, these situations would almost certainly have been no different.

There have been general threats made against federal agents, according to DHS. The agency claims to have uncovered bounties issued by Mexican cartels and others against officers. There have been bomb threats made against facilities in Texas and New York. There have been generalized online threats, too. When I wrote about the documented assaults for The Washington Post last year, a DHS official pointed to the arrest of Ray King, who’d made threats against ICE agents on social media … in part because he was angry that agents were obscuring their identities.

DHS posted a snippet of a voicemail apparently left for the agency, which it described as “DISGUSTING,” including “vile rhetoric” that was part of “violence and dehumanization” their agents faced.

The content of the snippet? The caller suggests that masked agents are “bad guys” and calls for them to be identified and subject to “public shaming.” You may be the judge of the extent to which this constitutes violence.

Another DISGUSTING voicemail left for our brave ICE law enforcement officers.

The vile rhetoric of sanctuary politicians comparing ICE day-in and day-out to the Nazi Gestapo, the Secret Police, and slave patrols has consequences.@ICEgov officers are facing more than 1,000%… pic.twitter.com/i1UR0WoQ1i

— Homeland Security (@DHSgov) October 15, 2025

There have been threats made against and harassment of specific officers. The spouse of an officer in Texas allegedly received a voicemail suggesting that her husband’s fate would mirror that of the Nazis. On Jan. 24, the day Alex Pretti was killed by federal agents, a caller allegedly suggested that an officer should kill himself because “all of your friends are fucking murderers.”

An attorney working for ICE had her personal information posted online. Agents in Portland were identified in flyers posted around the community. A similar incident in California led to a massive raid on a home, footage of which aired on Fox News. An ICE official had a small amount of garbage dumped on their lawn, with a sign reading “ICE IS TRASH.”

Three women were indicted in September for livestreaming while following an agent home and revealing his address. They shouted to neighbors that “ICE lives on your street and you should know” and encouraged anyone watching to “come on down.” Again, though, this was apparently a case in which the agent’s identity was determined because he had been at work when the women started following him.

These are isolated incidents, none of which resulted in violence. Again, most of the assaults against ICE agents have occurred while agents were conducting their duties, including when engaging with protesters. A central reason that assaults against ICE officers have spiked since Trump took office is that there are more officers engaged in more actions and conducting more arrests.

It is fair to note that ICE’s policy of allowing agents to shield their identities might have done exactly what they claim: protected them from off-duty retribution. But it is also fair to note that, even when agents have been identified outside of work, they have not been targeted with violence.

It is also fair to note that ICE has at times conflated “identification,” “harassment” and “violence” to suggest that even criticism and calls for accountability are unacceptable attacks on their employees.

If protection isn’t the reason for masking, you might ask, then what is? Well, perhaps what I theorized back in May (and at the beginning of this article): that masking allows ICE agents to avoid accountability for their actions. We’ve seen far more videos of ICE and Border Patrol agents manhandling protesters and immigrants than we’ve seen the reverse; we’ve seen incidents in which unidentified agents clearly engaged in unacceptable violence against protesters (drive-by pepper spraying, firing non-lethal munitions at peoples’ heads). Who, for example, killed Alex Pretti? We don’t know.

A judge in Minnesota suggested that ICE’s indifference to the law likely has meant that it “has likely violated more court orders in January 2026 than some federal agencies have violated in their entire existence.” If ICE views itself as above the law, which seems more likely: that its agents wear masks so that they are protected against the public — or so that they are protected against legal accountability?

Photo: Federal agents in Minnesota. (Chad Davis/Flickr)

On Wednesday, FBI investigators executed a search warrant targeting the headquarters of the Fulton County, Georgia elections department. The warrant, obtained by ProPublica, sought physical ballots and election records from the state’s 2020 election.

I happen to have just written about how President Trump and his allies are attempting to relitigate the 2020 outcome with an eye toward testing systems in advance of 2026 and 2028, so I won’t regurgitate that here. Instead what I will do is revisit one of the more frustrating periods of my professional life, when, on what seemed like a daily basis, I spent my time debunking conspiracy theories about the results of the election Trump lost.

Many of which were centered in Fulton County and its primary city, Atlanta.

Trump and his allies targeted Georgia with conspiracy theories incessantly, mostly because the margin in the state was narrow enough that tossing out 12,000 Joe Biden ballots could have handed Trump a victory. This isn’t speculation on my part; you may recall that Trump was recorded trying to browbeat Georgia’s secretary of state into flipping the Georgia results.

There was even a brief flurry of action in December 2020 when Trump ousted Attorney General Bill Barr (in part because Barr admitted there was no evidence of fraud) and entertained replacing him with Jeff Clark, who dutifully cobbled together an official letter from the government telling Georgia that its results were suspect. Cooler heads prevailed, at least until the first week of January.

This time around, there are no cooler heads running the Department of Justice or the FBI. So it is up to me, it seems, to remind everyone of how we can be so confident that there was no significant fraud in Georgia in the election five years ago.

Again, a lot of what you might hear about Fulton County has not only been debunked but was debunked even before Trump left office the first time around. The thing about a water main break? About poll workers pulling secret ballots out from under a table? Neither of those things was an example of fraud — or even particularly interesting. They were simply Unusual Things That Happened™, and so were tasked by Trump’s allies with the important work of Raising Suspicions™. That’s all Trump ever really wanted to do, get enough people to think, hmmm that’s odd, that they wouldn’t mind if Georgia’s election totals were thrown out.

More than five years ago, state officials refuted Trump’s claims about fraud in the state.

Georgia Secretary of State staff about to hold a press conference refuting 1 by 1 the claims President Trump made on the call with @GaSecofState pic.twitter.com/dFOehpxula

— Justin Gray (@JustinGrayWSB) January 4, 2021

It’s important to recognize that these lies were not victimless. Two women, Ruby Freeman and her daughter Shaye Moss, became the faces of alleged fraud in Fulton County (in no small part because they were Black women), including in a series of false claims from Trump and his attorney Rudy Giuliani. But the state cleared them of any wrongdoing and Giuliani withdrew his allegations and was ordered to pay more than $140 million in damages.

Fulton County, like every county in Georgia, used Dominion voting machines in the 2020 election. This meant that an automatic machine recount was completed fairly quickly. There was also an additional recount conducted by hand; neither changed the results significantly. And what those results show is that Fulton County did shift to the Democratic candidate between 2016 and 2020 — but so did 84 other of Georgia’s 159 counties.

In fact, Fulton County swung left less than most of the counties that surrounded it. It was a focus primarily because there were a lot of votes cast for Joe Biden there, not because there was an unusual number of votes cast for him.

On average, Georgia counties saw 5.6 percent more votes cast in 2020 than in 2016. In Fulton, the increase was only 2.2 percent, ranking it 102nd out of 159 in terms of turnout growth.

In other words, if Fulton County was a hotbed of fraud, it was amateur hour compared to all the other counties around it.

That Georgia used Dominion machines, though, meant that it (and Fulton County) got smeared with the conspiratorial idea that machines had been rigged to give the election to Biden. An analysis I completed in December 2020, though, showed that most swing-state counties that used Dominion machines still voted for Donald Trump.

In fact, more counties that didn’t use Dominion machines flipped from Trump to Biden than did counties where the machines were in use.

Over time, more allegations about Fulton County emerged. Tucker Carlson, then still at Fox News, elevated a claim that ballots in the county had been counted more than once, inflating Biden’s totals. Carlson told a tale of a warehouse alarm being tripped and a mysteriously open door — a story that ended up being about an employee accidentally being locked in the building after using the restroom. There’s no evidence votes were counted more than once, as evidenced by the fact that those votes were counted two more times in the recounts and the totals didn’t change to a significant degree.

Then there was Dinesh D’Souza’s movie “2000 Mules.” D’Souza claimed that rampant illegal voting had occurred when people hired to collect ballots stuffed them into drop boxes. From the moment the movie was released, it was obvious that D’Souza had no evidence of drop boxes being stuffed, much less this happening as part of a grand conspiracy, much less that there was a discernible number of votes cast, much less that those votes were illegal. In time, D’Souza too was forced to admit that his movie didn’t show what he claimed it did.

None of this is to say that Fulton County’s handling of 2020 was flawless. The slow counting that occurred kept the door open for Trump and his allies to suggest that something untoward was occurring. The man in charge of elections in the county was ousted in February 2021.

But administrative clumsiness is not fraud. It is not close to fraud. And nothing remotely close to fraud has been identified in five-plus years of feverish searching by the president and his allies.

So what do we make of the FBI search that unfolded on Wednesday? Well, we should look at it in the context of Trump’s past efforts to use claims of fraud to secure power despite election results. And we should recognize it as, if nothing else, an effort to degrade trust in elections systems. Remember when he was impeached for trying to get Ukraine to launch an investigation of Joe Biden before the 2020 contest? He and his team knew that the investigation itself would raise suspicions, even if nothing was ever proven (or even surfaced). Here, too, the investigation is itself part of the effort to shape perceptions.

Here’s what we can expect to happen next. It will probably be the case that the seized information will yield some new allegation, just as Trump and his allies gaining access to other federal information led them to make fresh, dubious claims about old, settled issues.

And we can assume that when no smoking gun of 2020 election fraud emerges, as it almost certainly won’t, Trump supporters won’t view that as a repudiation of their shared belief in wrongdoing. It will, instead, simply be treated as proof that the fraud was so significant that records were destroyed before Trump and FBI Director Kash Patel could get to them.

Leading them to advocate strongly for changes making it harder for “fraud” to occur — changes that will invariably make it harder for Democrats to win or to have their victories certified.

Did I forget any debunked Fulton County conspiracy theories? Email me.

Photo: Trump talks with Pence in the Oval Office before heading to Senate campaign rallies in Georgia, Jan. 4, 2021. (White House/Flickr)

Perhaps the most enlightening comment former special counsel Jack Smith offered during his hours-long testimony before the Republican-led House Judiciary Committee centered not on Trump but on what Smith was tasked with doing.

The case his team had assembled in obtaining an indictment against the then-former president, Smith said, was ”built to be tried in a courtroom, not in the media.” But here he was, facing questions from legislators not in a courtroom but on Capitol Hill — which, for all intents and purposes, means “in the media.”

There wasn’t much use to the hearing. There’s no actual question to adjudicate. No serious and unbiased observer questions Smith’s objectivity or credibility and no serious observer questions that Trump tried to overturn the 2020 election, triggering the riot that overwhelmed the building on Jan. 6, 2021. In effect, then, the existence of the hearing necessarily served to reinforce the falsehood that there was a debate in the first place.

Smith summarized the importance of recognizing reality in his opening remarks.

“The rule of law is not self-executing,” he said. That is, the bounds of the law are real only to the extent that they are respected. Smith, better than most, understands what it looks like when that respect evaporates.

The product of the hearing wasn’t a studious consideration of the validity of his work since, as stipulated above, there was nothing serious to mull over. Instead, the primary output of the hearing was probably a tidy stream of social-media-friendly video snippets. Members of the House (nearly all of whom will soon face primaries or reelection) saw an opportunity to make news and most of them tried.

What this means, in effect, is that the hearing not only didn’t resolve any tension between reality and surreality, it simply dug each side in a little deeper.

This wasn’t Smith’s fault. He very carefully responded to Republicans’ questions/attacks, offering a cautious deliberateness that will undoubtedly be read as hedging by the Republican media universe. But the decisions he and his team made about who to charge and when and with what were, again, centered on obtaining a conviction, not on muting Sean Hannity. He couldn’t help but leave plenty of space for Republicans to use cheap little rhetorical tactics aimed at collapsing trust in him and, by extension, his very-obviously-valid investigation.

Not to sound like the middle-aged man I am, but at one point in time the effort to convey reality would have been aided by the presence of an intermediary layer between the legislators and the public. Congress would hold a hearing and reporters would observe it and then they would tell the world what happened. They did so imperfectly, absolutely, and not without introducing some small measure of bias. But these, too, were largely accounted for and corrected through the systems in which they worked: being unfair or dishonest was a good way to find yourself looking for work.

What the public would have heard about the hearing wasn’t, say, Judiciary Chairman Jim Jordan’s (R-Ohio) hyperactive dishonesties about Smith and the aftermath of the 2020 election. It would (ideally at least) be a summary of the hearing and his actions leavened with an anchor of truth. That Smith’s findings are credible, as is he, and that Republican politicians are biased in a way that Smith clearly isn’t.

These days, Americans are generally unencumbered by any such anchor. If they want to just get the “JORDAN HUMILIATED BY DEMOCRATS!” take, it’s there on YouTube. If they only want to hear “DERANGED SMITH COMMITS PERJURY,” they can subscribe to any random Republican legislator’s social media feed. Or, for that matter, the president’s. There is no filtering layer between the event and the public; there’s no net to trap the lies. More accurately: There are plenty of ways to drink the raw sewage without waiting for the water treatment plant to do its work.

It’s important to recognize that the presentation of falsehoods from legislators (particularly Republican ones today) is not solely a function of dishonesty. There are some, like Jordan, who think it’s clever to shade the truth to get what they want. There are also a lot who simply haven’t done the homework and trust the Jordans of the world to tell them how it is or what to say. Republican Rep. Michael Baumgartner of Washington admitted during the hearing that he really didn’t know anything about Smith’s work, but then ended his time at the mic by equating the special counsel to a soldier in Vietnam who endorsed the massacre of a village. If we can’t even get members of Congress to take five seconds and figure out the difference between true and false, what luck are we going to have with the public? And that’s considering those who actually perform their duties in good faith.

While Smith was testifying, we learned that the White House had shared a manipulated photo of someone arrested by the Justice Department to show her crying, a new line crossed by an administration fully committed to prioritizing the scoring Online Points Against Their Enemies over dutifully informing the public even if they seek to persuade them. It’s just another form of lying, one for which a large segment of the media will make excuses — or offer cheers.

Consider Smith’s line in light of that: he was aiming to meet the careful standard of a courtroom, not the non-existent standard of rhetorical disputes in the mass media. These two things used to be in relative alignment, even if they differed in rigor and methodology. But they no longer are, because one is a traditional system of rules and self-respect while media has shattered into a ruthless marketplace dominated by snake-oil salesmen. And elected leaders consistently see more value in becoming snake-oil franchisees than in telling anyone what they don’t want to hear.

Photo: Capitol police officers embrace on Thursday, August 5, 2021, in the Rose Garden of the White House. (Flickr/White House/Adam Schultz)

Donald Trump has started a new members-only club that demands a high entry fee. But instead of being situated in southern Florida or the rolling hills of New Jersey, this one is virtual. And instead of adopting a gaudily romantic name, this one is just called “the Board of Peace.”

The president rolled out this new initiative in an event held during the World Economic Forum’s annual meeting in Davos, Switzerland. He’d already offered some public details — member countries were asked for $1 billion in contributions; its focus would be on rebuilding war-ravaged Gaza — but the Davos event was the official launch party. Representatives of several of the countries that had already signed up to participate were there, including United Arab Emirates, Hungary and Pakistan.

You will notice that those countries are not traditional allies of the United States. That’s telling, given that some are concerned that the Board of Peace might be intended to be a center of power that exists in opposition to the United Nations.

In fact, nearly all of the countries that agreed to work with Trump’s Board of Peace are ones that are rated relatively poorly on an important metric: adoption of democracy.

The Economist Intelligence Unit’s annual Democracy Index assesses the strength of each nation’s commitment to democracy (or lack thereof). In 2024, the average score across the world — on a zero-to-10, least-democratic-to-most scale — was 5.2. The score among the identified founding members of Trump’s Board? 4.5 points. That includes the U.S.; removing our country drops that average to 4.3.

The EIU’s analysis slots countries into one of four categories, ranging from authoritarian to full democracies. Nine of the countries that signed up for Trump’s Board fall into the “authoritarian” group. None — including the U.S. — are considered full democracies.

Given that the focus is ostensibly on Gaza, you might assume that the imbalance above is a function of regional partners, non-democratic Middle Eastern countries wanting to invest. But it isn’t. It’s a panoply of weak democracies and autocracies that span the globe.

Why does this matter? In part because the U.S. has traditionally sought to participate primarily with international actors that share our political values. Trump appears to be building a pay-for-play alliance that gives bad actors access to American power, perhaps with an eye toward aggregating power to contravene the existing convening international body, the United Nations.

Put another way, this matters because these are not historically the sorts of partners that America has sought out. But, then, Trump is not historically the type of president we elect.

Below, the trend in the EIU’s index by country over time.

Photo: Donald Trump at an event in December. (Flickr/The White House)

At the outset of his second term in office, Americans were willing to give Donald Trump the benefit of the doubt. Every incoming president has something of a honeymoon period after inauguration, a few weeks or months when people (generally meaning independents and the opposing party) haven’t yet decided that he is terrible.

Trump wasn’t really a new president, as such, so there was no guarantee that he’d experience the same lift. But he did. In his second month, despite his furious effort to immediately reshape the government and the country, Trump generally enjoyed net-positive approval on key policy issues.

No longer. By the end of last summer he was underwater across the board. Even on immigration, an issue that was perceived to be a particular strength of Trump’s, significantly more Americans now see him doing poorly rather than well.

(I’m not going to spend time here getting into the debate over whether Democrats ought to have focused on the economy or on immigration when attacking Trump because it’s pretty clearly been settled. Attacking strengths is more valuable than highlighting weaknesses because you have the potential of eliminating that strength. It’s really not that complicated.)

The categories used above are necessarily broad, since it’s tricky to repeatedly ask poll respondents their view on dozens of specific policies. But YouGov has been asking people about a subset of Trump’s immigration policy: deportation. Over the past year, he’s gone from being plus-18 on it to minus-8. That’s largely because of a surge in opposition from non-Republicans — but also a small drop among Republicans.

One of the most remarkable recent poll findings came from CNN. Respondents were asked if they saw Trump’s presidency as a success or as a failure. Most said “failure.”

In fact, Americans were more likely to describe Trump’s second term as a failure than they were to use that word to describe his first term in the immediate aftermath of Jan. 6, 2021. Back then, only half of independents said Trump’s first four years in office were a failure. Now, two-thirds of them say his second term has been.

But note those numbers among Republicans. Nine in 10 approve of his deportation policies, even now. He’s well above water on approval on the various policy issues and overall. And while 1 in 9 see his presidency as a failure — which seems like a lot! — 8 in 9 don’t.

Maybe opposition to what ICE is doing will prompt him to reconsider his approach. Maybe concerns about how Republicans might fare in the midterms will do it. But he himself is not running for reelection (however much he likes to daydream about doing so) and has demonstrated an indifference to opposition that borders on the pathological.

When Renee Good was shot and killed by an ICE agent earlier this month, Trump and his team were quick to portray her as a “terrorist” who tried to run the agent over, neither of which was true. But then Trump learned from a CBS News interview that Good’s father was a Trump supporter.

“When I learned her — her parents, and her father in particular is — I hope he still is, but, I don’t know, was a tremendous Trump fan, he was all for Trump, loved Trump,” Trump said. “And, you know, it’s terrible. I was told that by a lot of people. They said, oh, he loves you. He was a — I hope he still feels that way. I don’t know. It’s a hard situation. But her father was a tremendous — and parents were tremendous Trump fans.”

That’s what he took away from Renee Good’s killing: It was primarily bad because her father was a Trump supporter.

That is what he takes away from all other criticism, too. Unless you are or were a supporter, he doesn’t care.

Photo: Trump “dancing.” (Flickr/White House)

I’m a bit tired of regurgitating all of the times that private citizen Donald Trump disparaged then-President Barack Obama’s golf habit from 2009 to 2017. You can search for yourself and see; on the campaign trail in 2015 and 2016 it was a regular part of his patter. When Obamacare kicked in, Obama would probably be off golfing!!! Obama talks about climate change — but flies a jet to Hawaii to play golf!!!!! “You can’t be playing golf when Japan is burning and crashing”!!!! And so on.

At one point — as I have also pointed out repeatedly — Trump told 2016 voters that he would be too busy as president to play golf, a pledge that got thrown out the window almost as soon as he actually took office.

Today marks the end of Trump’s fifth (non-consecutive) year as president. Over that period, he has likely played golf 20 more times than did Obama, 6 percent more golf.

I say “likely” because (as I have also noted before) Trump almost exclusively plays at private clubs owned by his family’s private company and prevents the media from knowing what he’s up to. I’ve tracked his daily calendar for both terms in office, though, and I am confident that my total (excluding days when it’s raining, for example, or when he wasn’t at a club long enough for a round) is accurate. (As for Obama’s total, that comes from the late CBS reporter Mark Knoller.)

Trump’s total is driven by his playing golf more often during his second term. In his first term, he averaged a round of golf every 5.6 days. This time around, he’s averaging a round every 3.9 days.

You can track Trump’s golfing and visits to Trump properties at pbump.net/outofoffice.

That increase in frequency has his cumulative total nearing twice as often as Obama. Obama played an average of 42 days per year, about 1 in every 9 days. Trump’s averaging 71 days per year — essentially a round every five days.

If Trump continues at his second-term pace, he’ll end up having played on 375 days from Jan. 20, 2025 to Jan. 20, 2029 — meaning that he will have likely played golf on the equivalent of a full year of his term in office. He’ll also have exceeded Obama’s eight-year total in only four years.

Trump doesn’t always achieve the things he desires. I’m confident, though, that he’s got the level of commitment necessary to make this one come to fruition.

Photo: Trump on a Trump Organization golf course last year. (Flickr/White House)

 

Over the weekend, ICE agents repeatedly used the killing of Renee Good to threaten Minnesotans recording their activities. "Did you not learn from what just happened?" They thought it was a "lesson" and now they'd be left alone. Didn't work.

— David Bier (@davidjbier.bsky.social) 2026-01-12T14:53:10.422Z

   

UNARMED WOMAN: “Shame on you.”ICE GOON: “Have you not learned from the past few days?”UNARMED WOMAN: “Learned what? What’s our lesson?”ICE GOON: *VIOLENTLY SNATCHES HER PHONE*(H/T @Mollyploofkins)

— The Tennessee Holler (@thetnholler.bsky.social) 2026-01-10T17:35:43.714Z

   

"You guys need to stop obstructing us, that's why that lesbian b*tch is dead!..They literally said, have you not learned? This is why we killed that lesbian bitch!"-2 ICE protesters arrested for legally following ICE tell our @jtcestkowski.bsky.social ICE slandered Renee Nicole Good. MORE⬇️

— Status Coup News (@statuscoupnews.bsky.social) 2026-01-12T22:26:52.492Z

   

“The driver, the ICE agent that had pepper-sprayed into the vents of my car, said, ‘You guys gotta stop obstructing us. That’s what that lesbian bitch is dead.”

   

REMINDER. “To all ICE officers: You have federal immunity in the conduct of your duties. Anybody who lays a hand on you or tries to stop you or tries to obstruct you is committing a felony. You have immunity to perform your duties, and no one—no city official, no state official,… pic.twitter.com/xoWDjOctLe

— Homeland Security (@DHSgov) January 13, 2026

 

Photo: Trump speaking on Jan. 6, 2026. (Flickr/White House)

Vice President Vance joined the daily White House press briefing on Thursday in order to talk about a new Justice Department position focused on fraud. But the details of that announcement (think Ronald Reagan’s dishonest “welfare queens” claims aimed at immigrants and updated for the social-media age) got buried by reporters’ questions about the killing of Renee Good in Minneapolis the day prior.

At the top of the briefing, Press Secretary Karoline Leavitt had set the tone.

“The deadly incident that took place in Minnesota yesterday,” she said soon after walking to the lectern, “occurred as a result of a larger, sinister, left-wing movement that has spread across our country where our brave men and women of federal law enforcement are under organized attack.”

Vance similarly accused Good of being involved in an effort broader than herself. In short order, one of the reporters in the room called him on it.

“You just suggested that this woman who was killed was part of a ‘broader, left-wing network,'” the question began. But since it came from Fox News’s Peter Doocy, what followed wasn’t, “What’s your evidence for that.” It was, instead, “Who do you think is behind this ‘broader, left-wing network’?”

Vance tried to offer evidence anyway. Surely there was some undergirding organization, he said. After all, “when somebody throws a brick at an ICE agent or somebody tries to run over an ICE agent, who paid for the brick?”

A red brick costs 64 cents. From whose deep pockets was that fortune drawn?

Vance extended the idea to Good herself.

“How did she get there? How did she learn about this?” he said. “There’s an entire network — and frankly, some of the media are participating in it — that is trying to incite violence against our law enforcement officers.”

We’ve seen this before. When protests against Israel’s military campaign in Gaza emerged in 2024, the right scrambled to insist that there was some well-funded, interlacing organization behind it. (Then, the indescribably expensive thing protesters could somehow access was $20 tents.) The right often acts as though anyone who disagrees with their politics are brainwashed, deranged into action against their true interests by some nebulous conspiracy.

At this point, though, the charges from the administration are consistently narrow. It’s not just that there’s a “broad left-wing network.” It’s that grassroots organizing over the internet itself comprises that network. If you read a story that prompts you to want to protest ICE, if you see that activists are using whistles to alert community members to the presence of immigration officials, if you scramble to a local school to record immigration actions: Bad news. You’re part of this nefarious conspiracy.

In September, after the killing of Charlie Kirk, President Trump and his allies attempted to tie the shooting to an organized left-wing effort. Trump signed an executive order, National Security Presidential Memorandum 7 (NSPM-7), that directed federal agencies to uproot “sophisticated, organized campaigns of targeted intimidation, radicalization, threats, and violence” that ply their trade “through a variety of fora, including anonymous chat forums, in-person meetings, social media, and even educational institutions.”

NPSM-7 is titled “Countering Domestic Terrorism and Organized Political Violence” — precisely the allegations leveled against Good by Homeland Security Secretary Kristi Noem and others.

The ACLU and others have written about how this directive threatens free speech. We see how that works in the Good killing. Beyond the lies being told about Good’s intent, we see how Vance and Leavitt (and countless others online) loop her into a broad conspiracy simply by virtue of being in that place and apparently expressing opposition to ICE’s actions. Since her death, right-wing media has documented Good’s involvement in grassroots “ICE Watch” efforts. But this is effectively a post hoc rationalization for her killing — after Noem and others had already rushed to identify her as a domestic terrorist.

There is a difference between a community and a terrorist organization. There is a difference between gathering information online and being indoctrinated. There is a difference between seeking to get involved and being a conspirator.

We should not assume that the administration and its allies are blurring those differences because they misunderstand them. We should assume, instead, that they are blurring them in order to increase the cost of engaging in the innocuous activities of grass-roots organizing.

Photo: A protest outside the White House this week. (Geoff Livingston/Flickr)

On Thursday morning, multiple major news outlets published analyses of videos from the killing of Renee Good by federal officials in Minneapolis. The New York Times. The Washington Post. CNN. Each neatly dispels the narrative offered by the Trump administration immediately after the shooting, establishing clearly that Good was not trying to run down ICE agents and, in fact, does not appear to have even injured any as she sought to escape.

This was obvious soon after the shooting. The videos analyzed by the news outlets were available Wednesday afternoon, showing precisely what the annotated and professionally edited presentations do. But the standard of evidence used by news outlets is stricter than the one used by a layperson, so it took longer for them to formally reach the same, obvious conclusion. It took longer for them to dismiss the White House’s presentations as false than it did for external observers, in part because they have an obligation not to be wrong that isn’t shared by regular people or, it seems, the president.

A gray zone emerges. As media outlets push to confirm what happened, people with access to much of the same information have already reached conclusions. Bad-faith actors — again including Donald Trump — are able to make flat assertions about what occurred that their loyal and/or incurious supporters accept. Critics of the administration coalesce around their own, more accurate understanding of events, at times chiding the media for being so slow in joining them.

In the days when news was published in daily newspapers or precisely at 6 p.m., this was less of a problem. Not only because there was no expectation of immediacy but because there didn’t exist independent systems by which regular people could view the same information as professional news-gatherers. But now there is, and an awful lot of framing and debate occurs in the space between an event and its being cemented as objective fact.

It’s a bit like election results, an analogy that I draw intentionally. Votes are cast on Election Day and, often, in the weeks prior. In some places, vote-counting continues for hours or days, creating a similar gray area. Initial results might suggest one thing and final votes another; voters and politicians are urged to wait for the latter instead of making assumptions about the former.

Sometimes, though, those voters and politicians know that waiting is disadvantageous. In 2020, Trump knew that, as more votes were counted, it was likely to mean more votes for Joe Biden. There was a plan in place to take advantage of the gray zone — to insist not only that initial information was correct but that subsequent information (that is, more votes) were suspect. Not only did Trump attempt to establish an election result before vote-counting was done, he tried to position the finalized vote totals as fraudulent because they differed from the premature story he’d presented.

We see this following news events, as well. As mainstream media slowly figures out what happened, Trump and his allies (particularly but not exclusively) rally around a story that comports with their politics. As more information comes out, that information isn’t used to adjust their position; instead, it’s used to demonstrate the bias and untrustworthiness of the media. If the final conclusions of the press don’t comport with their story, it’s the press’s fault, not the story’s.

We’ll see how the Good shooting evolves over the next few days. Recent history, though, suggests that Trump and others who are (grotesquely) attempting to blame Good for her own death are much more likely to decide that reality is inaccurate than that they themselves are.

Photo: Trump at a press conference shortly after the 2020 presidential election. (White House/Flickr)

They will lie about you and then kill you and they will kill you and then lie about you. 

And I don’t mean some ambiguous, imaginary “they” like the one the right invented and blamed for trying to kill President Donald Trump. I mean an actual they, a they that includes federal officials and federal law enforcement and Donald Trump himself. 

The killing of Renee Good in Minneapolis on Wednesday was obviously preventable. Confronted by federal law enforcement officers on a snowy street, video recorded from the scene shows Good attempting to maneuver between a stopped car and an agent standing near her front bumper rather than face arrest. As the car moved forward, brushing the agent to the side, he drew his weapon and shot her. Good was killed. 

But Trump and Homeland Security Secretary Kristi Noem offered very different descriptions of Good’s actions. Noem called her a domestic terrorist and suggested that she’d targeted federal officers with her car. Official statements from DHS echoed these claims. Trump shared distant footage of the incident on the social media platform he owns, describing with words a scenario very different than what anyone watching videos from the shooting would have seen. 

Describing Good as part of a “Radical Left Movement of Violence and Hate” — she was a poet — the president claimed that she had “violently, willfully, and viciously ran over the ICE Officer, who seems to have shot her in self defense.” Based on the clip, he claimed, it was hard to believe the agent survived. Other footage, though, shows him walking to Good’s car and back to other federal vehicles before being driven from the scene. 

What’s striking about the administration’s response isn’t simply that it’s dishonest. A year into Trump’s second term and a year into this era of unshackled immigration officers, we should only be surprised if their initial presentations hold up under scrutiny. Instead, what’s remarkable about the immediate responses is that they include no statements of regret or sympathy. Hours after a federal agent shot and killed a young mother, the government couldn’t even be bothered to extend its condolences. This wouldn’t have made the effort to impugn Good’s life and actions any better, certainly, but it would at least indicate some sense of humanity. It would suggest that even critics of the administration live lives worth preserving and appreciating. 

This administration has no interest in offering that impression. Good, by virtue of being in a position of friction with federal agents, became an Other. And Others are not people for the administration and its allies to mourn. They are people to disparage and mock and accuse. 

The grand theory behind American democracy is that our elected officials hold power with the consent of the governed. This has led past presidents to seek to demonstrate their commitment to even-handedness, to treat those who voted for them and those who voted against them equally, as Americans. There’s an aspect of self-preservation to doing so: the president or his party might need those voters” support in the future. Usually, though, this approach was rooted in the idea that the presidency really was a position centered on the public as a whole. 

This is not how Trump governs. He tends to and cares about only his base, and only as long as his base is enthusiastic about him. He is not an American president but a MAGA president, someone who leads red states with magnanimity and leads blue states like the head of an occupying force. There’s no effort to build consent among his opponents. There is only an effort to redouble his support among his supporters. 

So he and his team and his allies lie. They take the tragic death of a woman who’d just dropped her 6 year-old off at school and turn it into an obviously false narrative about heroic federal agents fending off a deranged left-wing lunatic. And rest assured that they will continue to do so if need be, just as they decided that Kilmar Abrego Garcia’s life needed to be destroyed because they’d embarrassingly screwed up and deported him. 

The Trump administration empowered a federal officer to shoot a woman from close range, reportedly in the head, and then blamed the woman for her own death. They posthumously determined her to be an enemy of the state because they need her to be one in order to keep their supporters from questioning how Trump uses his power. 

They will lie about anything else they feel the need to lie about, too, and for the same reason.

Photo: A vigil in Minneapolis, Jan. 7, 2026. (Chad Davis/Flickr)

Over the summer, CNN included an unusual question in a national poll of Americans: How politically connected did they feel to others of the same gender, race or generation?

Most Americans replied that neither race nor gender were particularly relevant to their politics. Generation, though? Six in 10 said that they had a lot in common with others in the same age group — about the same percentages that said gender and race weren’t relevant to their politics.

There’s an interesting pattern here. Men are among those most likely to say that gender isn’t relevant to their politics. And White people are among those most likely to say that race isn’t relevant to their politics.

On seeing this, my first thought was a quote from sociologist Robert Terry that I’ve had stuck in my head for years (and that makes numerous appearances in my book): “To be white in America is not to have to think about it.” Men and White people hold disproportionate power in the U.S. and often (though certainly not always) reject the idea that gender and race play a role in their doing so. Their rejection of shared politics with peers might reflect a rejection of the idea that such commonality offers political utility.

How, then, do we explain the results by generation? Older people also hold disproportionate power in the U.S., in part thanks to the growing size of the over-65 population. (I guess I’ll once again point to my book on this subject.)

Well, something interesting happens if you break out the generational splits by party. Democrats (and Democrat-leaning independents) are more likely to reject generational commonality as they get older. Older Republicans (and leaners) are more likely to endorse commonality.

This undercuts my original thesis about the response being downstream from power. After all, an obvious explanation for the inverted trends by party is that young people are more heavily Democratic and old people more heavily Republican, meaning that young Democrats do have more in common with people of their generation than do older Democrats. The same holds for older Republicans, relative to younger ones. Older Republicans, meanwhile, are certainly not lacking for political power.

Maybe, then, what’s being tracked is perceived power. Maybe groups that feel less powerful (regardless of how accurate that is) perhaps see common political concerns with their peers.

If we look at gender and race by party, the picture doesn’t get much clearer.

Democratic (+leaning) men are more likely to say gender isn’t relevant to politics than Republican men and White Democrats are more likely to say that race isn’t relevant than White Republicans. If you squint, you can see how that comports with the revised thesis — the GOP is, at the moment, heavily focused on the false idea that White men have been targets of systematic repression — but it’s not terribly clear.

Which is probably the best place to leave it: Sometimes one set of numbers doesn’t answer all of the questions we might have.

Thanks to Ariel Edwards-Levy for sharing more-detailed breakdowns of the data.

Image: From my 2023 book, a map of where baby boomers lived in 1990 (dashed-line circles) and in 2020 (solid lines).

The report on President Trump’s health published by the Wall Street Journal on Thursday was not particularly reassuring. The president has shown increasing signs of aging, the Journal’s reporters noted — unsurprising given that he’s the oldest person ever to have been inaugurated to the position. He continues to eat poorly and to adhere primarily to his own medical advice, a dubious approach for someone coming up on his 80th birthday.

But, of course, Trump spun the report on Friday morning as a complete exoneration of any concerns. On the social media platform he owns he wrote that his doctors had asserted that he is in “perfect health” — an assertion made in the article not by his doctors but by Trump himself.

Trump also asserted that his mental acuity has been tested three times, and that he “ACED” the questions the cognitive exam posed. Rather than providing reassurance, though, one might wonder: Why is Trump’s mental fitness being tested so often?

The answer to that question will probably be unsatisfying for both Trump supporters and Trump critics. He isn’t being tested for cognitive impairment particularly frequently; he’s just talking about old tests a lot.

Here is a timeline of the three reported tests.

January 12, 2018

The first exam was conducted during his first term in office, back when his personal physician was still Dr. Ronny Jackson. During a press conference, Jackson informed reporters that Trump had been given an annual physical and, while delineating the elements of the examination, noted that Trump’s cognition had been screened.

“A cognitive screening exam using the Montreal Cognitive Assessment was normal, with a score of 30 over 30.”

This is not atypical, given that Trump was already in his 70s at the time. What was unusual was that Trump seized on this test as being evidence of his cognitive superiority — which, as I’ve written, is a bit like bragging about your superhuman physicality because your covid test was negative.

Sept. 13, 2023

The next cognitive exam was conducted during the Biden administration, after Trump had already declared his intention to seek reelection to the presidency.

pic.twitter.com/ibBepwOC2e

— Steven Cheung (@StevenCheung) November 20, 2023

Again, this is a normal examination for a person who is 77 years old. But it also allowed Trump to elevate his successful navigation of the screening process to suggest that Joe Biden ought to be public with his own results.

April 11, 2025

Trump’s most recent cognitive test was not, as his post on Friday might suggest, conducted recently. Instead, it was given to him after he’d returned to the White House, during a physical last April.

A report from Trump’s current physician detailed the test in sterile language.

Trump had a follow-up exam in October of last year, during which he received seasonal vaccines and in preparation for international travel. A report from the doctor after that examination does not suggest that it included a cognitive screening.

In the abstract, that makes sense. Such tests are currently recommended annually in the absence of any signs of cognitive impairment.

Whether insisting that a test passed nine months prior constitutes a recent measure of cognitive fitness should be considered a sign of impairment would, I suppose, be up to the doctor to judge.

Photo: Trump grins at Mar-a-Lago in December. (White House/Flickr)

During a tour at my son’s new school this summer, another father struck up a conversation. He asked about my son’s reading and described his efforts to get his child to pick up books more frequently. A key problem, he insisted, was that kids weren’t instructed as they had been when he and I were young: no more recitations and writing drills, among other things. He had hired an outside tutor in order to make sure that his child was learning the way he thought they should learn.

Perhaps it’s because I have friends who are teachers or perhaps it’s because I spent some time working in an elementary school, but his approach struck me as ridiculous. Teachers in public schools have spent years learning about teaching methodologies before they even get into a classroom and then hone their techniques as they teach. They necessarily have a much better sense of effective teaching methodologies than does some guy who hasn’t been in elementary school since he was a student during the Reagan administration.

This exchange has popped into my mind sporadically over the past few months, and I’ve come to believe that my response to his insistence wasn’t really a function of who I knew or my experience. It was, instead, a recognition that the world evolves and changes and that how things were is never necessarily better than how things are.

It takes a long time for people to recognize that the world isn’t static. Some probably never do. We are born into a world and learn how the world works as we age and then feel as though we have a handle on things. But the world, frustratingly, has kept on changing and evolving, until the world that we understood has new tweaks and gaps and systems that we don’t recognize or immediately understand.

That’s not what we don’t recognize, really; everyone knows that things change. What we don’t recognize for a long time, if ever, is that even the static world into which we were born was itself a product of change. The static reality that we had a firm grasp on by the age of 10 or 15 was someone else’s bizarre new way of existence that made no sense and was probably not all that sensible. We plunked into a constantly shifting timeline but only really recognize the shifts that diverge from the world we first learned about.

I suspect — and I’ll admit to probably being somewhat parochial here — that this understanding of having been plunked down into a roiling, churning reality probably comes more quickly to people who live in environments where such change is a constant. My wife and I took our kids down to Soho this morning, a Manhattan neighborhood near where she and I used to live. Walking around was a constant stream of “this used to be X” comments about buildings and storefronts. The places that hadn’t changed were the exception.

This is the nature of New York City. I imagine that it is much less the nature of, say, a small town in an agricultural region. I imagine that in such places, with fewer things to change and change happening at a less rapid pace, it’s harder to see change as the constant and easier to see it as anomalous.

I could spin this out into some observation about politics and American culture, but it’s Dec. 31 — about seven hours, as I write, from the new year. I’m writing this in part because the trip to Manhattan reminded me that things change but that the ship of Theseus is still the ship of Theseus. This year has been one of personal change extending far beyond my son’s new school. It is reassuring to remember, to recognize that the universe is chaos, not order. Change is eternal. And it is not necessarily change for the worse.

So trust yourself. And, for God’s sake, trust your kid’s teachers.

Photo: Goldberger’s Pharmacy on 1st Avenue, as it looked in 2016 (and photographed by me). Below, via Apple Maps, what it looks like today.

I’ll cut to the chase: No.

We as Americans do have the habit of being selective about how we describe the motivations and priorities of voters. Most people cast their ballots in support of candidates from the party with which they agree, rather than engaging in a complicated analysis and ranking of policy positions. (This, in fact, is a central reason that political parties exist!) When a candidate loses or wins, however, they — and their supporters — and their detractors! — often ascribe the victory or loss to specific things that they advocated or opposed. If they ran on putting billboards on the Washington Monument and were elected because their opponent was caught on video abusing a puppy, the now-inaugurated candidate will happily inform the public that they must put a Pepsi logo on the National Mall because it’s What The Voters Wanted.™

Over at Bluesky today, there was a little flurry of a debate over the extent to which supporters of Donald Trump’s 2024 candidacy were explicitly supportive of the sweeping deportation effort that’s now underway. One side of the debate argued that Trump (and his surrogates) spoke frequently about deporting millions of people; there had even been signs at the convention reading “Mass Deportation Now.” The other side noted that conventions are only rarely considered by the voting public, and that many people voted for Trump for reasons that had nothing to do with any promises about deportation. (Like, for example, inflation.)

Again, the latter argument is correct. And we know that because Trump voters themselves indicate that they oppose the sort of deportations that are underway.

Even before Trump took office in January, polling showed a wide gap between support for deporting any immigrant who was living in the country without proper authorization and deporting those who’d committed crimes. The specific numbers depended on the wording of the poll (see results from AP-NORC and the Wall Street Journal below), but the gap was consistent: Far more people endorsed narrower deportation efforts than broader ones.

Those numbers weren’t broken out by party, however. So let’s look at June polling from YouGov, which presented various deportation scenarios to respondents.

Again, we see that deportations of immigrants convicted of crimes enjoys wide support. But support declines under different circumstances. Only a quarter of Americans — and only half of people who say they voted for Trump last year! — support deporting immigrants who have kids that are citizens. Less than half of Trump voters support deporting people who’ve lived in the country for years without committing crimes.

I’ll note that this doesn’t mean that half of Trump voters oppose deporting people who’ve lived in the U.S. for years without committing crimes. In fact, only a third do; the other 20 percent say they aren’t sure if such people should be deported.

But it is clearly not the case that Trump voters supported sweeping deportation of immigrants regardless of circumstance back in June. From that, it follows that it is unlikely that most of them supported such a sweeping deportation strategy last November, when they were casting their ballots. Some did, yes. But it’s very unlikely that most did.

I’ll add one last thing. While I was still at The Washington Post, I occasionally wrote about Trump’s deportation policies and looked for images from the convention of people holding those “Mass Deportations Now” signs. As it turns out, this was not easy to do. They were not ubiquitous or constant; I believe (and am happy to be corrected) that they were trotted out for only one part of one night.

Meaning that while they came to represent Trump’s position in the eyes of his opponents, it’s probably not the case that they represented the views or understanding of most of his supporters.

Photo: Trump at Mar-a-Lago on Nov. 26, 2025. (White House/Flickr)

Every so often, I Google the names of my grandparents. The internet keeps expanding backward as more pre-internet information is digitized so it’s interesting to see if anything else has surfaced.

I wasn’t lucky enough to spend much of my adult life with them or wise enough to spend more of time with them when I could. Swimming through the internet allows me occasional glimpses of slices of their lives, from Census records documenting their existences to tiny vignettes that emerge solely through the twin coincidences of having at one point been documented and that documentation being digitized.

A few months ago, for example, I did a search for my grandfather, Elwood Glass. In the past, I’ve uncovered his name in advertisements related to his work recruiting employees at Sohio or as a member of the American Chemical Society. Did you know that local newspapers used to document people coming to town for visits? Well, they did.

My most recent search, though, uncovered something different: a mention from a 1979 Cornell University alumi magazine. Both he and my grandmother attended Cornell, graduating in 1938. And 41 years later, their classmates were treated to an unusual update on their lives.

A lovely family tradition is that of Flora (Daniel) and Elwood Glass ’38, described in a recent issue of Woman’s Day magazine. Each year, for more than 30 years, they have saved their Christmas tree, cut the trunk in small pieces, and written the year in pencil on the cut ends. On Christmas Eve, the family (five grown children and—at last report—six grandchildren) gathers in front of the fire and throws into it a piece from each year’s tree.
Each child puts on a piece from his birth year and those from special years such as college graduation, marriage, etc, and now the grandchildren put on a piece from their birth years as well. The children who celebrate Christmas in their own homes have also started this beautiful custom.

What struck me about this update wasn’t that it offered some new information about an interesting part of their lives. What struck me, instead, was that this is a tradition in which I still participate.

---

In my garage, I have ten chunks of wood, each a segment of a Christmas tree from each year of the past decade. When I’m taking our tree down after Christmas, I cut about eight inches off of the trunk, adding it to the collection. By the following Christmas Eve, it is seasoned and dry, so I write the year on the base with a marker — and then cut a slice off of it and each of its predecessors.

The night before Christmas is one of the few days each year on which we use our fireplace. I start a fire and my wife and (admittedly reluctant) kids sit around it. I have a file on my laptop that records the events of the years that correspond to each piece of wood — itself a modern update to the heavily scrawled upon piece of paper my grandfather used to use — and I read out what happened in that year. Or, at least, what we thought was important to include; the ceremony (such as it is) ends with our figuring out what events of the current year should be included on the list.

As the Cornell alumni magazine suggests, someone is picked to toss the piece of wood from each year into the fire. New jobs, the kids’ birth years, etc., are reserved for the person for whom the year was the most meaningful.

When we would do this at my grandparents’ house as a kid, there were decades worth of logs (or, in some cases, small slivers of wood) and often more than a dozen people vying for the title of most significant year. At our house these days, the lucky person is generally whichever of my kids wants to throw it in, since a lot of what’s documented occurred before they were born. But that’s the point: It’s a way of documenting and sharing family history that they otherwise don’t know.

Which brings me to the other interesting part of that 1979 update about my parents: That at some point in that time period, this family tradition had earned a write up in a national magazine.

---

My quiet hope was that this mention would turn out to have been some multipage spread, one in which a photographer from Woman’s Day came to my grandparents’ house in Cleveland Heights and took pictures of the process. I remember certain parts of that house distinctly: the green, high-pile carpet and yellow chairs in the living room; the pervasive, slightly acrid smell; the low-ceilinged attic stairs with the bookshelf of kids books on the landing. To be able to see a photo of that space, much less of my grandfather’s list (where did that end up?) would be remarkable.

But this meant tracking the article down. It meant, most likely, another attempt to see how much of the past the internet has already absorbed.

I was hampered by the vagueness of the reference. The alumni magazine was dated February 1979, suggesting that the Woman’s Day issue was probably during the preceding holiday season. At that point, the magazine came out monthly, and I tracked down the November 1978 and January 1979 relatively easily. The latter cost me $5 to access a downloadable PDF — a gamble, since it was unlikely that the magazine would be presenting Christmas tips in January (much less that the alumni magazine worked on such a relatively quick turnaround schedule). And, sure enough, the mention of my grandparents was in neither.

The most likely issue, of course, was the December issue, but I was having trouble locating a digital version of it. There was one accessible through Proquest, a clearinghouse of old periodicals that I could access from the New York City Public Library, but that meant carving out time to head into the city to do so. I set the project aside for a few weeks.

As it turns out, though, no such trip was necessary. Picking the idea back up earlier this month, I remembered that the NYPL allows people outside of the city to get library cards. I did so, allowing me to log into Proquest remotely.

The story about the fireplace logs wasn’t in the December 1978 issue. It was in the December 1977 one, featured in a section of household tips called “Neighbors” just across from a cigarette ad.

MEMORY TREES
Each year for over thirty years we have saved our Christmas tree, and after winter-mulching perennial plants and small shrubs with the lopped-off branches, my husband has cut the trunk into small pieces, each with the year written in pencil on the cut end. Then every Christmas Eve we gather in front of the fire and throw into it a piece from each year’s tree for all those thirty-odd successive years, remembering as we do so the major events of the year. Each child—now grown, of course—puts on the piece from his birth year (as the grandchildren are beginning to do too) as well as from years that held special events for him-college graduation, marriage and such. Thus each Christmas Eve becomes a time filled with happy memories of all the years we have lived in this house. Now the children who have Christmas in their own homes are starting the same custom for their children, and last year when for the first time we spent Christmas with a child’s family, they gave us part of their tree to take home and add to our collection so we could have an unbroken sequence.
Flora D. Glass, Cleveland Heights, Ohio

No photo of the family or of my grandparents’ house, but still a wonderful window into their lives. It’s very fitting that my grandmother would go into unnecessary detail about the handling of the old tree; she was an avid and careful gardener. Just reading the brief description allows me to visualize crowding around their fireplace on that green carpet, my mother, sister and me debating my cousins and their families about who deserved to throw in a particular sliver of log.

I would love to know how this came about. Not only the tradition, the origins of which are murky to me. (Was it something my grandparents themselves had inherited? Or was it something they invented, like saying “Asciugamani!” when toasting — Italian for “hand towel”?) But why did my grandmother send this in? Was she reading Woman’s Day when she spotted a call for family traditions? Perhaps she sat down to type it up at a typewriter in the small office at the top of the stairs — a room that I remember as being off-limits to the grandkids but that we lingered around anyway to review the framed family tree that hung on the wall just outside of it. I remember her and my grandfather as being fairly reserved, but this was obviously a notable enough achievement that it warranted sharing with the other alumni of Cornell University.

Holding the magazine in my hands, I had another thought, one that I’ve had often before: What would my grandparents think of me? What would they have said to me? What might they have told their friends? This is an under-appreciated part of what happens when we lose loved ones: We not only lose them but we lose the bond between them and us. We’re left imagining what present-day interactions might have been like — a longing that’s being exploited by those AI-grandparent apps.

This is a large part of why I was Googling my grandparents in the first place. It’s an effort to add onto the piecemeal memories I have of them with whatever other information exists out there. It’s an attempt to build as complete a picture as I can of people I didn’t really know well, besides what I picked up during visits … or as we were throwing logs into a fire.

Then, as I was holding the magazine, I realized that it was itself the endpoint of a connection between my grandmother and me. I, like her, find enough value in the tradition to share it with others, as I have on social media each year since the pandemic. Without knowing that she’d once done so, I sat at my own keyboard in my own office and sent out a message to the world about a thing that our family did that was important to me. Across a half century and halfway across the country, the same impulse from grandmother and grandson.

And now this essay about it, which will have to substitute for the alumni magazine.

A family tradition: pieces of Christmas trees from past years that are burned as we talk about the highs and lows of those years. pic.twitter.com/2cKmG35GsY

— Philip Bump (@pbump) December 24, 2021

Header photo: The segments of tree trunk that we used for our 2024 tradition.

Americans went to bed on Saturday night with news that a shooter had killed two people at Brown University and woke up on Sunday morning to news that gunmen had massacred 16 people on a beach in Australia. The terror and disgrace of gun violence, hours apart on the opposite sides of the world.

The 16 deaths on Bondi Beach, where two men targeted Jewish people celebrating the start of Hanukkah, were the worst such incident in that country this century. There have been 10 mass-killing incidents using firearms since 2001; more people were killed and wounded on Sunday than in the other nine mass-killing incidents combined.

Here, I’m using a definition of “mass killing” that includes incidents in which at least three people were shot to death, excluding the shooter. This definition is certainly debatable, but it’s what I’m using because it’s also what’s used for Mother Jones’s database of shooting incidents in the U.S.

If we add mass-killing incidents in the U.S. since 2001, you can see that the Australian incidents are quickly dwarfed. (You can mouseover or click the colors to see each country’s incidents.) Nine of the 120-plus incidents in the U.S. since 2001 had larger death tolls than Bondi Beach. Three resulted in more injuries.

You might have noticed that the chart above doesn’t include the deadliest mass shooting in U.S. history: the attack on a concert in Las Vegas during Donald Trump’s first term in office. That’s because it obliterates the scale of the chart on both axes.

What happened at Bondi Beach was horrible. It would also have been only the tenth deadliest mass shooting in the U.S. this century.

It’s worth noting that the U.S. also has a significantly larger population than Australia. But U.S. mass killing incidents have been deadlier even relative to population. The incidents in Australia killed 1.86 people for every million current residents. The incidents in the U.S. killed 2.12.

In fact, the shooting at Brown University was relatively unremarkable in the context of American gun violence. According to the Gun Violence Archive, there have been 22 incidents in which more people were killed in a mass shooting (defined as an incident in which at least four people are shot) and 24 incidents in which more people were wounded this year alone.

The incident at Bondi Beach was only the second-worst mass killing in Australian history. The first was a shooting rampage in 1996 that left 35 people dead. In the wake of that incident, the Australian government implemented a number of strict limits on firearms possession, helping to tamp down incidents like the one this weekend.

Following its more numerous massacres, the U.S. has not done the same.

Photo: An AR-15 at a range in Connecticut. (National Archives)

It remains the case that a good way to learn how people feel about things is to ask them. So you don’t have to simply assume, say, that people are hankering to work up a sweat before hopping onto a six-hour flight or that they think the central failure of the airline industry can be summarized as “sweatpants.” You can just contact a bunch of people over the phone and online and ask them to tell you what it is that they are concerned about.

Which is what YouGov did. And what they found is probably not surprising: The things that people find most annoying about flying are prices, delays and discomfort.

In fact, more than 6 in 10 Americans pointed to ticket prices as a major problem with flying. Half said the same of cramped seats, delays, hidden fees, and staffing shortages (which, of course, lead to delays). And waaaaaaaaaaay at the bottom of the list came “passengers dressing too casually,” which only 8 percent of respondents described as a major problem.

I’ve highlighted it above because it’s useful to contrast that issue (such as it is) with other highlighted ones, like delays and fees. Delays and hidden fees are very much issues over which the administration and the Transportation Department have some influence — yet Transportation Secretary Sean Duffy has instead repeatedly cajoled people into spiffing up before jumping in their C-group Southwest seat to Bozeman.

On most of these issues, by the way, there isn’t much difference between Democrats and Republicans. When looking only at those with an opinion, we see that the biggest gap in opinion between Democrats and Republicans is on the availability of ground transportation at airports, which is perhaps related to the fact that residents of big cities (where traffic is often more of a problem) are more likely to be Democrats.

The next two biggest partisan gaps are on a lack of aircraft safety — where Democrats, critical of Duffy, see a bigger problem — and on this issue of attire, where Republicans are more likely to offer complaint. (Just after that comes a probably correlated partisan divide, with Democrats being more likely to complain that dress codes are already too strict.)

The takeaway here, though, is that less than half of Republicans see attire as even a minor problem. More than half say that prices are a major problem, including a majority who identify hidden fees as a major problem with flying.

I’m afraid I have some bad news for them on each of those fronts.

Photo: What flying looked like in the 1950s — at least in a British Overseas Airways ad.

Last night, instead of going to bed slightly earlier than usual, I decided that I would instead watch (yet again) the 2006 Alfonso Cuarón film “Children of Men.”

This is a film that evokes strong feelings. It is at best bleak; its ultimate triumph (spoiler alert) sees the protagonist floating through the fog in a small boat. The plot centers on the collapse of humanity and it is resolved with “well, maybe not”? It is not a fun movie, but it does seem fitting for a dark Sunday evening.

So, anyway, that’s what I put on. As it started, I was surprised to realize that it was set in an increasingly approaching future: the once distant year of 2027. Filmmakers are both advantaged and disadvantaged by being unable to completely predict the future. It means that, once that future arrives, we see how far off the mark their predictions might have been. That deviation from reality, though, also means that films set in what was once the future still seem futuristic once that future becomes the past.

When I recently rewatched the original “The Running Man,” a few months ago, I was similarly surprised that the “future” had come and gone. Released in 1987, it was set three decades later — meaning nearly a decade ago. While some aspects of that movie matched the reality of 2017 life (enthusiasm for reality TV foremost among them), it was otherwise … a bit off the mark.

So I assigned myself a project: Find movies set in the future and compare their release dates to the years in which they were set. Which futures are now past? Which are looming? Which aren’t even close to arriving?

The chore of figuring out which movies might qualify and the data on setting and release dates was made infinitely easier by the existence of a Wikipedia page documenting movies set in the future. Some of the included films are dubious (“Home Alone”?) but it was useful for what I had in mind.

So, below: When future-looking movies were released and set, ranked by setting date. I pared the Wikipedia list down significantly, using a few criteria:

• No superhero movies, because who cares.
• The plot had to depend to some extent on the film being set in the future.
• The setting had to be a specific year, not a range or multiple years.
• It had to be a movie I’ve actually seen.
• Only one movie per franchise.
• If I wanted to ignore any of those rules, I could.

Here is the result.

It’s fun to imagine that these movies are all set in the same future. That the train in “Snowpiercer” is also trying to outrun the cyborgs from “The Terminator.” That RoboCop was abandoned so D.C. instead turned to the precogs in “The Minority Report.” That “Avatar” and “Elysium” document different parts of human space exploration in the same year.

There are actually nearly a dozen films on the Wikipedia list that are set in or around 2025. I included only “Her” because of the criteria above — and because it is one of the rare future-set films that actually came close to predicting the actual year in which it was set. (That it was released relatively shortly before that future probably helps.)

“Children of Men” will not prove as prescient, mostly because there have been children born since 2009. The element of the film that centers on the British government rounding up and detaining immigrants, though? Perhaps a bit closer to the mark.

Photo: Promotional still for “Children of Men”. (Universal via IMDB)

The always informative Aaron Reichlin-Melnick brought the Deportation Data Project site to my attention. It does what you would think, providing aggregated data on detentions and deportations undertaken by the federal government.

Obtained directly from the government, the information is useful in part because it is broken down into a number of different categories. One can see, for example, the prior criminal histories (if any) of those arrested by ICE. Data are also broken down by state, date and disposition; if there’s information provided by the government, it’s indexed and available for perusal.

Granted, it’s a bit hard to peruse in its raw, tabular form. So I created a quick interactive, displaying monthly totals by state and category. Want to see arrests by criminal history in Delaware? Want to see the percentage of arrests in D.C. that targeted people with criminal histories? Go for it.

Following Reichlin-Melnick’s lead, I also broke out the data so that you can look at all arrests in a month or just those of people who weren’t already in law enforcement custody. (ICE will often arrest and deport people who were already detained by local law enforcement agencies.) You can play with the tool below.

So far this year, there have been five special elections to the House of Representatives. Republicans have won three of those five races — not surprising, since they also won those three seats in last year’s general election.

What is surprising is that, in each of those five races, the Democratic candidate saw double-digit improvement on the final margin over both last year’s Democratic candidate and how Kamala Harris did in the district. On average, the races shifted 16 points to the Democratic House candidate and 17 points relative to Harris.

Those presidential results come from the essential site The Downballot, which has also tracked special election outcomes. Across those contests in 2025 (meaning, including state-level races), the average shift to the left has been about the same: 13 points.

The most recent contest, as you are probably aware, was in Tennessee’s 7th District on Tuesday. Again, the Republican Party held the seat, but much more narrowly than it or Donald Trump won it last year.

Final vote totals aren’t in yet, with the New York Times estimating that about 95 percent of the vote is in (as of the time I compiled this data). But it’s already clear that turnout in the Tennessee special was relatively high for a House special this year — and the Democrat still improved on last year’s margins by 13 percentage points.

It’s important to reinforce that there aren’t a lot of lessons to be drawn from a set of five elections. Everything that’s available reinforces the idea that Democrats will do well in next year’s midterms, certainly, but we should be cautious about using five unusual contests to assume that Democrats will overperform the GOP by 10 points next November.

But this is also why considerations of turnout are useful. One of the reasons that special elections are an imperfect predictor of future outcomes, of course, is that they usually involve fewer voters. If an election instead includes most or all of the vote total for a November contest, we might assume it’s a better indicator of how that district is likely to vote. There are a lot of squishy words there — all or most, better, etc. — but the idea holds.

In each of the five special House elections this year, the Democratic candidate received fewer votes than the candidate in last year’s general, just as the Republican candidate received fewer votes than last year’s Republican. But in each case, the 2025 Democratic candidates captured a larger share of their party’s 2024 House vote totals than Republican candidates did of theirs. On average, Democratic special election candidates got 52 percent of the 2024 Democratic candidate; the Republican special election candidates got 37 percent of the 2024 Republican.

In Tennessee, Democrat Aftyn Behn got 7 votes for every 10 the 2024 candidate for that seat got. Her opponent, Matt Van Epps, got about 5 votes for every 10 the Republican got last year.

There are a lot of reasons such a discrepancy might exist. Perhaps a lot of people who voted for the Republican last year backed the Democrat this year. Perhaps Democratic voters were more energized to turn out to vote. Perhaps it’s a function of the difference between last year’s specific candidates and this year’s.

Whatever the reason, the dual pattern — less drop-off for the Democrat and a shift to the left in the overall margin — are working against Trump’s party. Even in a race where Republicans managed to turn out some 100,000 voters for a special House election (as will likely be the case in Tennessee), those patterns hold true.

Comparing House races to House races seems like a better point of comparison than contrasting the House special elections with the 2024 presidential contest. But, if you’re curious, here’s that comparison.

In this case, the shift to the left in Tennessee was more modest than in the other four special elections. Republicans investing heavily in the idea that this means that higher relative turnout in 2026 will further erode the leftward shift in House races might scroll up a bit and notice that in Arizona’s 7th, where turnout was relatively modest compared to last year’s House race, the shift to the Democrats was about the same as in the relatively higher turnout in Tennessee.

We risk overreading all of this, of course. This is what happens, though: in the absence of concrete information about next year’s midterms, an election of enormous significance, we dig through whatever tea leaves litter the bottoms of our cups.

Above, you can see five leaves in two different cups. Have fun with your forecasting.

Photo: Tracking the results of the 1954 election. (National Archives)

Somewhat buried in the pre-Thanksgiving conversation, Dartmouth College’s Brendan Nyhan offered an interesting observation in an essay for the New York Times.

“[T]he scale of the protests” targeting Donald Trump this year, he wrote, “is still not as large as one might expect, given the severity of the threat. During President Trump’s first term, millions of people protested when the situation was far less dire.” He offered one reason for that decline: “the lack of young people.”

Nyhan pointed to YouGov polling that showed less interest in October’s “No Kings” protests among younger Americans than older ones. He contrasted that with the protests that followed the killing of George Floyd in May 2020; then, younger Americans were more likely to tell YouGov that they’d participated in a protest.

I’ll note that this specific comparison is partly a function of interest. In 2020, YouGov polling shows that far more young people expressed agreement with the Black Lives Matter movement than did older people. This time around, older people — who are more likely to have lived through the Cold War struggle between democracy and autocracy — say they are more dissatisfied with the state of American democracy, the issue at the heart of the “No Kings” movement.

In his essay, though, Nyhan also points to another factor: a sense of hopelessness.

“The more persuasive explanation for the relative lack of young people in the anti-Trump, pro-democracy movement is that they are demobilized and demoralized. But it would be a mistake to blame them for this attitude. Older generations should instead recognize that the world we have created does not seem to offer a viable path to making change.”

This is almost certainly true. When Gallup reported that record numbers of young women expressed an interest in leaving the U.S., I noted that this was likely in part a function of feeling as though there was no means by which their frustrations could be addressed. Democracy depends on the idea that power jostles back and forth between interests. If you feel as though the system has excluded or marginalized you indefinitely, what’s the point of sticking with it?

Such a sentiment isn’t only detectable in protest apathy or an expressed (though generally not manifested) interest in emigrating. It’s visible in the prevalence of independent voter registrations among young people, in skepticism about the media as an institution and in the quick collapse of both Trump’s and Joe Biden’s approval numbers among the youngest Americans.

It’s also suggested by the broader contraction of young people’s engagement in politics.

I’m not referring to voting. Young people consistently vote less heavily than older Americans, for a number of largely structural reasons. (For example, young people often hold jobs that afford them less flexibility to cast a ballot and move more often, meaning that they need to re-register to vote and perhaps figure out where to do so.) But young voters were more likely to vote in 2024 than they were in 2008, once considered a high-water mark for young-voter participation.

But they voted more heavily still in 2020. And even with relatively higher voting rates in 2024, young people made up much less of the electorate than they do the adult population.

Again, though, that’s not what I’m talking about. Instead, I’m talking about other engagement with the political system.

The Cooperative Election Study is a national poll conducted around each federal election. It includes an evolving set of questions measuring how often respondents engaged in non-voting political activity: going to meetings, putting up a sign, participating in a protest, donating to a candidate.

On many of those metrics, young people — and older people! — indicated less participation in 2024 than they had in 2020 or prior years.

There were only two activities in which young people were more likely to report participating than older people in the 2024 survey: attending political meetings and attending protests. The percentage reporting participation in protests was down nearly 7 percentage points from 2020, the biggest drop among any of the three age groups. Meanwhile, the percentage of younger people who reported having engaged in none of the identified actions was up more between 2020 and 2024 than was the case among older respondents.

It’s worth noting that this was also true just among young people who identify as Democrats or Democrat-leaning independents. In 2020, only 44 percent of them indicated they’d engaged in none of the identified political actions (compared to 60 percent of young Republicans and Republican leaners). In 2024, more than half of young Democrats indicated that they hadn’t engaged in any of the activities (while young Republicans held steady at 60 percent).

It has been a disheartening political moment for young Americans (who are more likely, for example, to oppose the government’s approach to the war in Gaza), for young Democrats (who watched a Democratic president about whom they were often apathetic be replaced by Trump) and for young women in particular (given the revocation of Roe v. Wade, among other things). The response from older Democrats has often been something like “well, you should vote!” which isn’t wrong, as such, but is still clearly not going to be seen as sufficient.

There’s another interesting aspect to the numbers here that is worth pointing out. Remember: young Americans were least likely to express dissatisfaction with democracy in the U.S. in YouGov’s recent polling … perhaps in part because they don’t expect any better. Maybe that will change after the 2026 election, a point at which — should democracy actually be functioning normally — power is likely to jostle back toward the left.

Photo: A “No Kings” protester in October. (~jar{} on Flickr under Creative Commons license)