colinbmacinnes

This article was originally published by Quanta Magazine.

Thousands of miles from home in the steamy Amazon rainforest in the mid-1800s, the British naturalist Henry Walter Bates had a problem. More than one, really: There were thumb-size biting insects, the ever-present threat of malaria, venomous snakes, and mold and mildew that threatened to overtake his precious specimens before they could be shipped back to England. But the nagging scientific problem that bothered him involved butterflies.

Bates had noticed that some of the brightly colored Heliconius butterflies in the forest didn’t flit about like the rest; they moved more slowly. When he captured them and examined them under his makeshift microscope, he discovered that they weren’t really Heliconius at all, but astonishing look-alikes.

By the time Bates’s discovery reached the scientific cognoscenti in England, Charles Darwin’s then-new proposal of natural selection could explain why this brilliant mimicry occurred. Birds and other predators avoid Heliconius butterflies because they are toxic to eat, with a bitter taste. The mimics were not toxic, but because they looked so much like the foul-tasting Heliconius, they were less likely to be eaten. The closer the resemblance, the more potent the protection.

What Bates and many later evolutionary biologists couldn’t explain was how this mimicry was possible. Getting the right shades of aquamarine and fiery orange in the right places on the wings required a constellation of precisely tuned genes. Those traits would have to be inherited with perfect fidelity, generation after generation, to preserve the Heliconius disguise. Maybe real Heliconius butterflies could afford to deviate a bit in coloration because their toxins could still teach predators to stay away in the future, but the mimics needed to be consistently flawless replicas. Yet the random reshuffling and remixing of traits in sexual reproduction should have quickly disrupted the essential coloring patterns.

[Read: Scientists can now repaint butterfly wings]

Today we know that in many species the answer is supergenes—stretches of DNA that lock several genes together into a single inheritable unit. “They’re kind of a wild card,” says Marte Sodeland, a molecular ecologist at the University of Agder in Norway. This aggregated form of inheritance “has obvious advantages, because it allows rapid adaptation, but there’s a lot we don’t know yet.”

Supergenes once seemed like an evolutionary oddity, but the rise of genetic sequencing has shown that they are far more common than researchers believed. Not all supergenes may serve a function, but work in just the past few years has revealed that traits in a wide range of animal and plant species might be driven by these groups of genes that function like a single gene. Supergenes help wild sunflowers adapt to a range of environments, such as sand dunes, coastal plains, and barrier islands. In other families of plants, they produce subtle but important variations in their sexual organs and fertility that help prevent inbreeding. Research published last spring showed that in some fire-ant species, supergenes determine which type of social organization predominates—whether a colony has a single breeding queen or more than one. (Specific supergenes in humans haven’t been confirmed, but likely candidates have been found.)

Supergenes also seem to hold explanations for many long-standing mysteries of evolution, such as how species can sometimes adapt to new environments rapidly, how populations can sometimes evolve in different directions even while living close together, and why some species have “balanced lethal systems” of breeding, such that they must have two different versions of a chromosome to survive.

But supergenes aren’t all-powerful. Recent work on the evolution of supergenes is painting a nuanced picture of their effects. These theoretical models and studies of real populations have shown that supergenes often accumulate harmful mutations far more rapidly than other pieces of DNA do, and this can gradually lead to degenerative effects that undermine the original benefits.

---

The definition of a supergene is rather technical, and scientists still argue about its finer points even though the concept has been around since the 1930s. But at its simplest level, says Simon Martin, an evolutionary biologist at the University of Edinburgh, a supergene is a group of genes that are inherited together as a unit, often with a lot of other noncoding DNA.

“You can continue to produce two distinct traits with multiple genes and not worry about them becoming jumbled up,” Martin says.

That jumbling tends to occur during the production of egg cells and sperm. In that process, the maternal and paternal copies of chromosomes line up and randomly swap segments of DNA in a ballet called “recombination.” Recombination hedges nature’s bets about the value of different permutations of genes; it boosts genetic diversity and helps weed out harmful mutations.

The superpower of supergenes is that they block this. Typically, supergenes contain DNA deletions, insertions, or inversions (sequences that were cut out and spliced in backwards). As a result, those parts of the chromosomal DNA don’t align with a partner and are far less likely to recombine.

In the 1970s, researchers showed that this same mechanism—with misalignments in chromosomes blocking recombination in segments of chromosomes that then continue to lose genes—led to the evolution of Y sex chromosomes from X chromosomes in mammals. Sex chromosomes are essentially supergenes run amok. Both supergenes and sex chromosomes exist because there’s sometimes a benefit to having some sets of genes inherited together, says Deborah Charlesworth, one of the evolutionary geneticists who pioneered the sex-chromosome studies and who recently retired from the University of Edinburgh. In those cases, “it would be ideal to not have any recombination but to have the things that go well together stuck together for good,” she says.

To understand why that might be advantageous, think about doing laundry, says Emma Berdan, an evolutionary biologist at the University of Gothenburg in Sweden. Say you have a basket of white towels and a basket of red towels. Recombination does the equivalent of tossing both loads into the same drum, flipping on the hot water, and pressing start. What results is a bunch of pink towels. But the evolutionary equivalent of pink towels often isn’t a problem, Berdan says: A blending of traits can be beneficial.

Sometimes, though, life benefits from keeping its genetic laundry separated. For Bates’s Heliconius butterfly mimics, having a mix of color splashes from different genes could be disastrous. The butterflies reap the reward of mimicry only if they look enough like Heliconius to fool predators.

That’s why many researchers have been probing how supergenes arise and what the consequences for species might be as their supergenes continue to evolve. Understanding the origin of a supergene is “one of the most challenging questions,” says Tanja Slotte, an evolutionary geneticist at Stockholm University who studies supergenes in plants. “And it’s not a given that it’s even always possible.”

In one recent effort, Katie Lotterhos, an evolutionary marine biologist at Northeastern University, built a computer model to study the first tentative steps taken on the path from inversion to supergene. Her model, published in a special August issue of Philosophical Transactions of the Royal Society B on supergenes, showed that the larger the initial DNA flip-flop, the more likely a supergene was to evolve. The reason was simple: A larger inverted fragment of DNA was more likely to capture multiple genes and lock them together as a single entity. Any beneficial mutations arising within the inversion could then promote its spread as a supergene.

[Read: The shellfish gene]

But the more important insight from Lotterhos’s model was that inversions themselves do not necessarily provide an evolutionary advantage. If a suite of genes is already well adapted to its surroundings, locking it into an inversion will not suddenly allow it to take off as a supergene. That fact may help explain why complex vital traits aren’t routinely secured as supergenes: Ordinary selection pressures are often sufficient to preserve the traits.

The question of whether an adaptation precedes an inversion or vice versa, Lotterhos realized, might never be answerable. “What comes first, the inversion or the adaptation?” she says. “It’s probably a little bit of both.”

---

Supergenes offer robust advantages in the inheritance of adaptive traits, but they come at a cost.

Think back to Berdan’s laundry analogy: Washing red and white towels in a single load does eliminate the color differences between the two sets of linens. However, if you rip or stain a pink towel, you have an identical pink towel you can use as a backup. If one copy of a chromosome picks up a harmful mutation that breaks a gene, a functioning backup copy is likely to be on the matching chromosome to help the organism survive. And since recombination ensures that the mutation is inherited independently of other genes, natural selection can weed out the mutation over time.

For supergenes, however, that isn’t true. Because they rarely recombine, any harmful mutations they acquire tend to stay in place. The benefits of supergenes, then, could be accompanied by significant disadvantages. For example, Berdan and Benjamin Wielstra, of the Institute of Biology Leiden, have found that in the salamander called the crested newt, half of the eggs it lays aren’t viable because of all the mutations that have built up in one supergene. Its supergenes seem to be holding back its reproductive success.

Supergenes can also complicate the process of mating. In some species, supergenes create a breeding system that effectively has four sexes. Because of a supergene in the North American white-throated sparrows, for example, there are two “morphs” with dissimilar coloration and behaviors. Not only do males have to find females, but they must find a partner from the opposing morph. Otherwise, offspring will likely die, either from inheriting supergenes from both parents or from inheriting none. Only chicks that receive a “balanced lethal” inheritance of one supergene and one ordinary segment of chromosome survive.

With such a steep price, it’s a wonder that supergenes evolved at all, Berdan says. “Any set of variants is going to be really hard to maintain, especially over millions of generations,” she says. “That’s one of the big mysteries of supergenes.” She suggests that multiple types of selection might be working together to preserve supergenes, and that certain environments might be most conducive to their persistence in the population.

Ironically, one of the mechanisms that can sometimes preserve supergenes seems to be recombination—the phenomenon that they normally resist. Amanda Larracuente, an evolutionary geneticist at the University of Rochester, and her co-authors described such a case last April in eLife.

Larracuente wasn’t initially interested in supergenes or their evolutionary costs. Her focus was on selfish genes, segments of DNA that proliferate in populations without benefiting their hosts. She was fascinated by a selfish gene called Segregation Distorter (SD) that arose in certain fruit flies in Zambia and skews the sex ratio of their offspring. “It’s a sperm killer,” she explains, but it kills only sperm that don’t carry a chromosome with SD.

Sometime within the past 3,000 years, one version of SD ensnared a large piece of chromosomal DNA, creating a supergene known as SD-Mal that spread to fruit-fly populations throughout Africa. “It’s really the ultimate selfish gene,” Larracuente says.

DNA sequencing and analysis by Larracuente, the biologist Daven Presgraves, and their colleagues showed that chromosomes with SD-Mal accumulate harmful mutations, as predicted by the near-complete lack of recombination between SD-Mal and its sister chromosome. But the researchers didn’t find as many mutations as they expected.

The reason, they discovered, is that occasionally a fly will inherit two chromosomes with SD-Mal—and those two supergenes are just similar enough to allow some recombination between them. That recombination in turn makes it possible for a few harmful mutations to be purged from the flies’ supergenes over time.

“As it turns out, just a little bit of recombination is enough,” Larracuente says. She and Presgraves are now looking for other SD supergenes in wild fruit-fly populations for clues to the evolution and impacts of supergenes more generally.

Their results show that the purifying effects of recombination on genomes never cease to be important. The complex traits that the stable, predictable inheritance of supergenes make possible may be invaluable in helping species adapt, but even the supergenes can benefit from mixing things up once in a while.

When I was a kid, my dad did something on family vacations that perplexes me to this day: He ran. Every day, at least four or five miles, rising before the sun and before anyone else was awake. He wasn’t training for anything. He wasn’t trying to lose weight. There was no specific goal, no endpoint, no particular reason he couldn’t take the week off while in the greater Disney World metropolitan area, which, in July, is hotter than the surface of the sun. He was just running, like he had basically every day since time immemorial. My dad will turn 75 next week, and whenever you’re reading this, he has probably already been out for a run today.

My dad didn’t always run. He started a few years before I was born. One day he wasn’t a jogger, and the next day he was, even if he didn’t know at the time that the change was indeed permanent. When I asked him why he started all those years ago, there was no great motivator, no epiphany. “It was the cool thing to do at the time for people like me,” he said of the 1980s fitness boom. Now, when my dad doesn’t run, “I feel like I’ve lost something,” he told me.

For a lot of people, this is, without exaggeration, the dream: You decide you’d like to start doing something, you get past the initial phase of this new activity being hard and bad and a huge bummer, and then you do that thing for 40 years. It’s a deceptively simple fantasy—and, so often, an impossible one. Right now, I work out once or twice a week, which is less than I’d like to. I’ve tried to form various exercise habits over the years—I bought the equipment, I made a plan, I got out there and did it—and never quite reached the automatic stage, even though I observed it at close range for my entire childhood. My experience is extremely common among people who want to change how they do all kinds of things: to waste less money, to floss, to quit smoking, to drink less, to learn a new language. And it’s a salient dynamic at the beginning of a new year, resolution or not. New beginnings are seductive, and so is our own capacity for change. You tell yourself that, this time, you’re really going to do God-knows-what differently, but all too soon you’re reminded that forever is a pretty long time to keep it up.

Stories like my dad’s often serve as pop-psychological proof that you, too, could become a runner, if you really wanted to. But we all want things—human longing knows no bounds—and plenty of people do genuinely throw themselves into trying something new, without much success at converting those behaviors into lasting habits. If some people can just get up one day and decide to behave differently for the rest of their life, why do most people fail at it again and again?

The conventional wisdom on changing habits goes something like this: You can change if you really want to. Americans especially tend to see ourselves and one another as individuals with identical reservoirs of willpower, which some people choose to use and others do not. If you can’t figure out how to get up at 4:30 in the morning to make sure you get five miles in before commuting to work every day, which my dad somehow did for 30 years, you’re not trying hard enough, or you don’t want it badly enough, or you’re not motivated enough. Try harder.

[Read: The false promise of morning routines]

Now, as anyone who has ever tried anything might suspect, it sure seems like that idea might be bullshit. Or at least, many researchers have concluded that it does not account for an enormous amount of observed human behavior, according to the psychologist Wendy Wood. In her book Good Habit, Bad Habit, she explains that for the latter half of the 20th century, psychological scholarship more or less affirmed its righteousness. Attitude leads behavior, the theory went, and the circumstances in which you exist aren’t that important to the choices you make. Individuals do certain things and not others mostly because of their own conscious decisions; your fate is largely in your own hands. You can see how this logic permeated culture: The self-help and diet industries boomed, the government slashed the social safety net’s tires, the 1970s became the Me Decade.

In the past 20 years, the field’s tune has begun to change. According to Michael Inzlicht, a social psychologist at the University of Toronto, the most recent research suggests that conscious decision making plays a much more minor role in people’s actions than previously thought, and long-term behavior patterns largely aren’t created by stringing together a series of conscious choices. How people use the phrase self-control, he told me, tends to conflate two different things: a largely immutable element of someone’s personality (a trait) and a way that someone chooses to behave at a particular time (a state). Trait self-control varies from person to person, he said, and the amount you get is probably determined by some combination of heredity, culture, and environment. A person who has high trait self-control might be preternaturally punctual, whereas the timeliness track record of an average person—even one constantly trying hard to make it to things on time—might be more subject to the whims of circumstance.

The key distinction here, Inzlicht told me, is that a person who appears highly self-controlled to others—who is displaying a high level of trait self-control—probably isn’t exercising their behavioral self-control as much as you do. “People who have high trait self-control, they don’t actually engage in more restraint of their behavior and thoughts and emotions in the moment,” he said. Instead, they just aren’t tempted or distracted or diverted from their purpose as often or as effectively as the rest of us. For the small number of people at the far end of the trait spectrum, the things that others have to exercise self-control to resist every single time—sleeping in, skipping the gym, making impulsive purchases, having a cigarette even though they’re trying to quit—often just don’t get recognized as viable options in the same way that they do by the rest of us. This doesn’t mean it’s always easy for these people to sit down to study or to get out the door to exercise, but broadly speaking, they’re less readily pulled away from their plans by the siren song of novelty or opportunity, so they don’t have to rely on their active sense of restraint, with its far less reliable results, as frequently.

My father, I suspect, is one of the people at the far end of this personality continuum. It’s not just the running: He smoked for more than 20 years—back when everyone smoked, he would want me to stipulate—but he quit on his first try and never looked back. After he decided he liked running every morning, he added on an evening strength-training routine several days a week, and has stuck to the two-a-days for decades. He always eats breakfast, and the meal—a big bowl of Raisin Bran and a buttered, toasted bagel for much of my childhood—goes unchanged for years at a time. He is always reading a book, often about history, and has probably consumed hundreds of thousands of pages of dry descriptions of obscure military battles in my lifetime, for fun. He is what happens if you make the whole plane out of New Year’s resolutions.

Inzlicht described the evidence that any given person can increase their level of trait self-control to be more like my dad as “extremely weak,” but said that researchers should nonetheless still keep looking for ways it might be possible. Having the inherent ability to more easily form good habits and jettison harmful ones is enormously beneficial—people who can do that tend to be healthier, happier, and more financially stable. My dad is an easygoing, curious, nonjudgmental guy, and a very good dad. He’s not constantly working against himself to do the “right” thing, or to do anything at all. This is just how he is. “I don’t get out in that lane, I don’t get out in this lane, I stay in my lane,” he said, when I asked him about his routines. “These are the things that I like to do.”

[Read: The problem with being perfect]

According to Wood, the Good Habit, Bad Habit author, forming new long-term behavioral patterns is possible to some extent for most people, and it’s largely a function of learning to do something so automatically that you perform the task without having to consciously decide to do it, like brushing your teeth before you go to bed. She runs the University of Southern California’s Habit Lab, where she studies how and why people learn to change their behavior. She says that people who go to the gym a lot, for instance, don’t have to decide to go every time—they just sort of find themselves headed in that direction at the appropriate moment.

For those to whom habit-formation doesn’t come so naturally, the circumstances you’re in can make a big difference. Stability, for instance, is an enormous boon: Many people who leave work at the same time every day are able to rely on their routine as a cue to tell themselves that it’s time to go to the gym. If only half your workdays end when the gym is open, converting that choice to a habit can be much harder. Having money to buy the tools that make a new behavior easier or more rewarding is also enormously helpful, as is consistent access to the environs in which new tasks can best be performed. My dad started running on safe, low-traffic streets, which gave him the opportunity to realize that he really, really liked the rush of endorphins that is often called a runner’s high, which reinforced the creation of his new habit in ways that Wood has found to be crucial.

One study Wood described found that people who lived within four miles of a gym went much more frequently than people who lived farther away, even when the difference between the two groups was only a mile or two. Another study found that adding farm stands outside of schools and community centers in a low-income area of Austin, Texas, meant local residents ate more vegetables, even if nothing else was done to encourage people to change their dietary habits, or even to tell them the farm stands were there. “That’s part of the health advantage of higher-income folks,” Wood told me. “They live in environments that are more conducive to exercise; they’re less likely to live in food deserts; they have access to restaurants beyond just fast food.” For many people who make something out of their good intentions and healthy tendencies, those successes have been supported by policy choices that they had nothing to do with. To Wood, the implication is clear: If you want people to behave differently en masse, you’re going to have to change—to improve—the circumstances in which a lot of them live.

I do not mean to sound fatalistic here. It’s not that personal change or self-improvement is impossible—most people can change their habits and create new ones, according to Wood, if they set realistic goals and they’re able to create cues and rewards that effectively encourage repetition. Much of that involves tinkering with the circumstances of your existence that you can affect. For instance, I became a more frequent flosser by taking the package of floss out of my medicine cabinet and sitting it next to my toothbrush, where I could always see it. I used to procrastinate on washing dishes, but now I do them every day like clockwork, thanks to a Bluetooth speaker that I use to listen to podcasts while I stand at the sink. Having a clean kitchen, in turn, means I cook more—an activity I really enjoy—and resort to expensive takeout orders less frequently. I figured out what was stopping me from doing some of the things I knew I could do, and I tried to eliminate the obstacles I could control, to reasonable success. Figuring out how to do something a little less or a little more is likely to yield the best results for most people, even if it’s not going to turn you into a different human.

Before you do any of this, though, or before you decide you’ve failed, it’s probably worth making peace with who you are as a person. My irregular exercise habits don’t really bother me anymore, mostly because I do not take myself as seriously as I used to. I figure that I am who I am, give or take a reasonable capacity for marginal change. I have exercise equipment in my apartment that I could use more often, but I simply do not feel like it. I have never once felt like it, even if I have often wanted to be a person who does. What I can actually do for myself over the next year to make my life better probably will not include a spontaneous dedication to daily exercise. It may include more careful attention to, say, reading or cooking—things that I already love, and that are good for me. Before I left my parents’ house over the holidays, Dad made sure to pass on a few books he thought I’d enjoy. None of them is about war.

The New York Times exposé of President Donald Trump’s taxes is chock-full of tantalizing revelations. According to the report, the president avoided paying federal income tax in 10 of the 15 years preceding his election. In 2016, he paid $750 in federal income tax, less than one night’s stay in a suite at the Trump International Hotel in Washington, D.C.

But the story also left some big questions unanswered. How exactly has a self-declared billionaire avoided federal taxation for the better part of this century? Do these records prove that he’s poor, or astonishingly good at hiding how rich he is? And what’s the deal with all the money he’s losing on golf courses?

When I talked with several tax experts recently, including some with direct knowledge of Trump’s past tax records, three somewhat overlapping interpretations emerged: incompetence, malfeasance, and criminality. I wrote most of this before the president fell ill. It feels awkward to focus on this now, but on the eve of an election, the American people need to understand the questions swirling around the president’s finances. I hope the president fully recovers, and provides us with some answers.

Explanation 1: Trump’s businesses are doing terribly

One rather literal interpretation of Trump’s tax records is that Trump is a pretty bad businessman, full stop.

[David Frum: Trump just lost control of the game]

Since 2005, Trump has made hundreds of millions of dollars from The Apprentice and associated branding deals. He’s poured much of it into risky investments that have ostensibly failed, and he’s used those losses, on top of business deductions, to drive down his taxable income. In this way, he’s marshaled business failures to zero out his income taxes.

But the scale of Trump’s losses still stunned the tax experts I interviewed. “His losses are staggering, especially for somebody in real estate, for whom the tax code is very generous,” said Steve Rosenthal, a senior fellow at the Urban-Brookings Tax Policy Center, who has studied Trump’s tax returns from the 1990s. “In part, this reflects the changing nature of Trump’s business operations. He used to be a pure real-estate mogul. But in the last 15 years, he’s made more money as a television star who licenses his name to various products. He’s been incredibly unsuccessful at owning and operating stuff in that time.”

Trump is no stranger to losing money. Tax forms obtained by The New York Times showed that, from 1985 to 1994, Trump lost “more money than nearly any other individual American taxpayer.” But his net worth is estimated at $2.5 billion. He’s clearly very rich! So how has a clearly very rich person avoided paying income taxes for the better part of this century?

Explanation 2: Extremely questionable tax maneuvering

Trump claims business deductions like Louis XIV furnished Versailles: lavishly, laughably, with monarchical prerogative. For instance, Trump deducted $70,000 in hair stylists. “The tax law is clear that you don’t get to claim a business deduction for expenses that are fundamentally personal in nature,” said Ari Glogower, a tax-policy professor at the Ohio State University. “It’s the same principle that says you can’t deduct fancy suits just because you work in a nice office.”

More egregious, Trump seems to have paid $700,000 in “consulting fees” to his daughter, Ivanka—a sum he also claimed as a business deduction. “There could be a variety of games going on there,” Glogower said. “It could be about avoiding estate taxes, or avoiding payroll taxes, or avoiding state taxes. We need more facts to know for sure, but shifting income to your family through unearned consulting fees is a classic tax game, which Fred Trump used as well.”

[David A. Graham: Trump has nothing left up his sleeve]

Those examples, although garish, are small potatoes compared with Trump’s treatment of debt. Across decades, Trump honed a simple, devious strategy. He borrowed lots of money, lost it on failed investments, and took huge deductions on interest, depreciation, and operating expenses, which drove down his federal income tax. Very often, he would declare bankruptcy or otherwise renegotiate the terms of the original loan to avoid paying it back in full.

What’s so wrong—or deviously clever—about all that? Well, Trump’s trick, according to Steve Rosenthal of the Tax Policy Center, is all about what he did after the bank wrote down his debt. “So Trump would borrow money from banks, spend it all, and go back to his creditors and find a way to avoid paying the whole loan back,” said Rosenthal. “But when a bank writes off your debt, the discharged debt is income. You have to report it as taxable income. Trump never did.”

Instead, Trump repeatedly ducked his obligations to lenders and the IRS, perfecting the same playbook—borrow, spend, deduct, and walk away. Eventually, American lenders caught on to the gambit and stopped giving him money, forcing him to seek other sources of revenue, such as Deutsche Bank and international plutocrats.

And that’s where, just maybe, things took a criminal turn.

Explanation 3: A mystery only prosecutors can unwind

Starting about 10 years ago, the so-called King of Debt tried something new: spending his own money. He spent up to $400 million buying golf courses, hotels, and wineries. He also took on $400 million in personally guaranteed loans, exposing him to personal bankruptcy in the (very common!) event that he can’t pay back the loans in full.

These decisions are unusual for private real-estate companies, which typically take advantage of debt through the tax code and funnel their largest investments through partnerships that protect individual investors from personal bankruptcy. What’s more, it’s not clear how Trump suddenly came up with all of that cash. Eric Trump told The Washington Post that his father “had incredible cash flow and built incredible wealth….He didn’t need to think about borrowing for every transaction. We invested in ourselves.” But The New Yorker’s Adam Davidson wrote in 2018 that “The portfolio of assets that Trump owns does not suggest that he would have so much money that he can casually spend a few hundred million on a whim.”

It gets even weirder. All those golf properties look like horrendous businesses, having declared losses of more than $315 million since 2000, according to Trump’s tax records. In short, Trump is pouring a mysteriously large amount of money into opaque businesses suffering mysteriously high losses.

None of the tax experts I talked with said: “Yep, that’s clearly an illegal scheme.” But several told me that when companies, such as golf courses or restaurants, that seem to be losing lots of money nevertheless continue in operation for many years, prosecutors might become interested in investigating money laundering.

[Sarah Chayes: This is how kleptocracies work]

What is money laundering, exactly? Say I’ve got $10 million in illegally earned cash. What am I going to do with it? Buy a condo? Uh, no, the IRS will sniff that one out. So I have to find a way to hide this cash; to mix my dirty money with some clean money; to launder it. So I give my $10 million to a clever guy I know. I don’t want to buy anything from my pal directly. I want him to buy something else with my $10 million that produces a steady flow of clean money, such as a golf course that brings in $2 million a year in revenue. Every December, my pal takes $100,000 of that revenue for himself and sends me $900,000; the golf course keeps only half its income. Voila! I’ve turned my $10 million of dirty, un-spendable money into $900,000 of clean, beautiful cash flowing into my account, every year. Also, because the golf course is sending out 50 cents on every earned dollar to a couple of crooks, it’s going to look like a terrible business.

If Trump’s golf courses and other properties were serving as money-laundering enterprises for (let’s say) Russian oligarchs, we would expect to see three things: mountains of cash materializing from basically nowhere, several large money-losing enterprises, and a disinclination to abandon those failing properties. We’re seeing all of that.

Brian Galle, a tax professor at Georgetown University, told me that my three explanations aren’t mutually exclusive. “There is a relationship between your interpretations one, two, and three,” he said. “If you’re a naturally successful businessperson, there is less pressure to maintain your wealth with tax-evasive maneuvers, or to engage in full-fledged money laundering.”

Trump’s tax records are a clue, not a complete answer. Their incompleteness might be the most concerning thing about them. “As a citizen, my reaction to the Times report is: The president owes a ton of money, and we don’t know who he owes it to,” Galle said. “Many of his creditors are effectively shells, and it’s opaque who the parties and interests are. Maybe there shouldn’t be a freak-out, because he owes it all to Deutsche Bank. And they’ve never been involved in money laundering, right? Oh, that’s right.”

With two months to go before Britain is due to exit the European Union, the country is mired in political dysfunction; its political leaders agree on little, if anything; and the terms on which Britain will leave the EU are yet to be agreed on.

Now, then, seems like the perfect time for Boris Johnson to up the stakes even further.

The prime minister went to the queen today to request that Parliament be suspended, a rare move that tightens the political calendar, reducing the number of days lawmakers have to debate Brexit and other measures. The House of Commons returns from its summer recess next week, and Johnson has effectively dared his opponents to unite against him in a matter of days—or cede control of the Brexit timetable to him.

Johnson says he is willing to see Britain leave the EU without a deal entirely, a scenario that his government’s own analysis states would lead to interruptions in food and medicine supply chains. His opponents have raised the possibility of a temporary national-unity government to postpone the October 31 deadline for Britain to withdraw from the bloc. Both are willing to face another election, Britain’s third in four years.

How might the coming weeks play out? Here are three of the main scenarios.

No Confidence Leading to Caretakers

Among the main options available to anti-Brexit lawmakers and the opposition is to pass a vote of no confidence in Johnson’s government. Were the prime minister to lose, then the parliamentary arithmetic would be vital to deciding what happens.

Under the Fixed-Term Parliaments Act, which outlines the government’s next steps if a no-confidence-vote motion is passed, the government would grind to a halt for 14 days, during which it would fall on the leader of the opposition—in this case, Labour Party Leader Jeremy Corbyn—to try to form a government.

This wouldn’t be easy. Corbyn would need the support of a host of smaller parties such as the Liberal Democrats, the Scottish National Party, the Green Party, and the Welsh nationalist party Plaid Cymru—all of which, like Labour, oppose Britain leaving the EU without a deal. None of these groups has ruled out establishing a temporary unity government to prevent a no-deal exit by requesting more time to negotiate.

But not all of them are keen on putting Corbyn, who is popular with his party’s base but attracts skepticism among the wider public, in charge. In a letter to the Labour leader this week, Liberal Democrat Leader Jo Swinson warned that putting Corbyn in Downing Street would alienate lawmakers who oppose a no-deal Brexit, and urged him to consider nominating an alternative leader, such as the longtime Conservative lawmaker Ken Clarke or the longtime Labour lawmaker Harriet Harman. Since the Fixed-Term Parliaments Act doesn’t stipulate who should lead a temporary government—only that they be able to command confidence in the House of Commons—this would be possible. (If 14 days pass and no government of national unity is formed, it then falls back to Johnson to call for an election.)

A government of national unity—even an explicitly temporary one—is extremely rare. “The last time it was done was in 1940, when there was a coalition formed in order to fight the Second World War,” Tim Bale, a politics professor at Queen Mary University of London, told us. Whether the prospects of preventing a no-deal Brexit can overcome party politics is still unclear.

A Fall Election

In a second scenario, either Parliament or the courts manage to bind Johnson’s hands, preventing him from pursuing a no-deal Brexit. Unable to chart his own course, the prime minister could, then, trigger an election.

Lawmakers have limited time and few constitutional means to do this, though.

They could, for example, pass a law that would bind the government to ensure that it does not exit the EU without a deal and, if a deal cannot be agreed on, give Parliament another opportunity to decide what to do next. Under the British constitution—unwritten, but codified in various documents and through centuries of precedent—all government ministers are bound by the law. The prime minister is not above the law.

The speaker of the House of Commons, John Bercow, might help them by granting an emergency debate. He has consistently been friendly to rank-and-file politicians who want more say over the executive. (Bercow has already called the suspension of Parliament “a constitutional outrage.”)

In the three years since Britain voted to leave the EU, though, Parliament has consistently failed to wrest control of the Brexit process from the government. Pro-European members of the ruling Conservative Party have been reluctant to vote against their own leadership, and are wary of allying with Corbyn. Previous efforts to limit the government’s room to maneuver, even largely symbolic ones, have barely passed the House of Commons, with one being dismissed by Downing Street as “inconvenient rather than significant.”

If Parliament cannot block a no-deal exit, then who can? Former Conservative Prime Minister John Major has suggested mounting a legal challenge to stop Johnson suspending Parliament. The government is depicting its plans as routine, though, making it hard to argue against them. The Labour member of Parliament Ian Murray is pursuing a separate legal action in Scotland, which operates a distinct judicial system from the rest of the United Kingdom.

If any of these challenges succeed, the question becomes whether Johnson is able to pass a deal of his own. He might hope to bring home concessions from a meeting of EU heads of state two weeks before the Brexit deadline—and frighten previous refuseniks into voting for his deal as an alternative to a chaotic exit.

But if he still cannot pass a deal and a no-deal exit has been blocked, then he faces a choice: Ask for another extension from the EU or call for an election. He has already rejected the first option, saying that the October date is “do or die.” (However, it would not be the first time he has reneged on a promise.) The second option would require a vote of two-thirds of the House of Commons, and would almost certainly pass.

The timing then becomes crucial. A short election campaign takes six weeks, meaning a vote in the House of Commons next week would see polling day land around October 17. Parliament would not sit during this period, preventing it from blocking  a no-deal exit during that time. Johnson could go to the electorate—fed up with years of debate and discussion about Brexit—and present himself as the leader of the only party committed to leaving the EU at all costs.

Brexit Happens, Do or Die

What happens if both of those efforts—a vote of no confidence and bids to limit the government through legislation or in the courts—fail?

In effect, the government would not face any parliamentary scrutiny until mid-October, when the suspension of Parliament ends, by which time Johnson will be just two weeks away from his goal of taking Britain out of the EU—too little time for an election, and barely any time for last-ditch attempts to remove him from office.

Even if his opponents held out the hope of trying for a vote of no confidence, Johnson would have a little more than five weeks, without the hassle of explaining himself to Parliament, to negotiate a new withdrawal deal with the EU. Should he show any signs of progress in doing so, anti-Brexit Conservative MPs who would have considered defecting or voting against his government may feel it is their duty not to vote him down when the House of Commons returns.

Under Johnson’s plan, he will attend a final EU summit before Brexit on October 17 and 18 before putting any revised deal—if there is one—to Parliament on October 21 and 22. While lawmakers will have some room to attempt to remove Johnson from office before October 31, the die may be cast by this point and Johnson will be presenting MPs with a choice of deal or no deal. Either way, Brexit will go ahead on October 31 and Johnson will remain prime minister.

In a capital consumed with crime and punishment, Robert Mueller, a.k.a. Bobby Three Sticks, keeps the president in his sights.