Why are we so bad at judging risk?

On the night of November 10, 1975, Captain Ernest McSorley did something unusual for him. He slowed down to let another ship catch up. As John U. Bacon tells it, McSorley was known for aggressive decisions. This one was cautious and rational, and it may have cost the Edmund Fitzgerald the hour it needed. The ship went down in Lake Superior with all twenty-nine men aboard.

The Great Lakes aren't gentle. Their storms build steep waves close enough together that a seven-hundred-foot freighter can ride two crests at once, in water shallow enough for the bottom to strike a hull. In another Great Lakes wreck, a captain turned down an immediate rescue to try to save his ship, and lost ten men.

Resist the urge to name one cause. Weather, cargo weight, structural fatigue, a delayed decision: none was enough alone. But one part should make you angry. Bacon reports that over just a few years, regulators let the ship's required freeboard, the margin between waterline and deck, shrink by more than three feet. Nobody had to say out loud that safety was being traded for cargo.

The Fitzgerald's margin didn't disappear in one decision. It was traded away a little at a time, by people who never had to say out loud what they were giving up.

In rural Alabama, a young man bled to death while the ambulance took over an hour to cover six miles. Stories like his are how Malcolm Gladwell gets at a problem most Americans misdiagnose.

Americans think they have a problem with crime. Gladwell's argument is that they don't. They have a problem with violence, and most of it is expressive rather than instrumental. It isn't people plotting to break the law. It's people settling arguments.

That distinction changes everything downstream. If murder is mostly arguments that escalate, then having more police around matters, and Gladwell notes that American police departments are small compared to Europe's. Programs that teach young men to slow down in a moment of conflict start to look like real public safety, not soft extras.

The most surprising story is about measurement. Gladwell points to Memphis, which cut its homicide rate in half between 1935 and 1985. His argument is that violence didn't fall; medicine got better at saving people. Once you know that, you'll trust crime statistics a lot less. Definitions drift, reporting varies from place to place, and Gladwell describes a study of who died at the scene and who died on the way to the hospital in which, he argues, race shows up in ways that are hard to explain away.

The Second Amendment chapters are where I'd push back. Gladwell treats Supreme Court justices like fifth graders over grammar, and that kind of contempt hardens people on the other side. And just because an idea's origins are poison doesn't mean the idea is poison.

Gladwell follows stories through a lens he thinks is right, so leave room for him to be wrong. But the core distinction holds up, and it's about risk: we fear the stranger committing a crime, when the bigger danger is an argument that escalates. Misjudge which risk you face, and you'll pick the wrong fix.

Here's a number that should change how you think about the stock market. A study of U.S. stocks from 1954 to 1996, which Robert Shiller cites, found that only about two-thirds of the swings in individual stock prices came from real information about future profits. The other third came from changes in investor attitudes. If you assumed markets were nearly perfectly efficient, that's a lot of mood.

Shiller won the Nobel Prize in 2013, sharing it with Eugene Fama, the father of the efficient-markets idea Shiller spent his career pushing against. That alone tells you how live the argument is.

Not every idea here is finished. His proposal for government bonds tied to GDP is mind-opening in theory, but you should wonder whether it would hold under stress, because protection like that tends to fail exactly when you need it most.

The blind spot is simplicity. Good financial ideas need to be simple to be adopted, not just correct. Religion works on people because it's simple: a few rules, a story you can hold onto. Most of what Shiller proposes doesn't have that, and it's no coincidence so little of it has been implemented.

Two smaller ideas stick. Shiller's passage on forged paintings, worth far less than genuine ones even when they look the same, explains why AI-generated writing feels hollow: nobody wants something as valuable as real thinking to come effortlessly. And accountants, he points out, exist to preserve an organization's memory.

About a third of a price is mood. The people who keep score, and the institutions that remember, are what keep the mood from running the market.

In 1764, Robert Clive heard news of a battle in Bengal that almost nobody in London knew about yet. He mortgaged his property and bought East India Company stock. He was right. The Company had won, and within a year the Mughal emperor handed it the right to collect the taxes of Bengal.

That is how a trading company became a government. The Muscovy Company, chartered in 1553, had pioneered the joint-stock form: an entity with infinite life that people put money into, sitting beside a government that people put their rights and their time into. In India the line between the two dissolved. By Dalrymple's account, the Company raised an army larger than the Crown's. Members of Parliament held its stock while voting on its affairs.

It is also a story about division. Mughal rulers and rival powers came close to pushing the Company out more than once, and kept turning on each other instead. A country without a shared identity is easy to take apart, a lesson that runs from Yugoslavia to Ukraine.

Dalrymple ends by arguing that empire isn't disappearing, it's turning into global corporate power. That's where I part ways with him. Immerwahr makes the case that technology made owning land far less valuable, and they can't both be fully true. What I'll concede is that legal inventions like the joint-stock company might still allow a kind of pilfering under the cover of capitalism.

So who really ran India? Not the emperor on his throne, and not Parliament, at least at first. It was whoever controlled the tax collection and the army, and for a while that was a company accountable mainly to its shareholders.

In 1945, an American soldier fighting through Manila didn't know the Philippines was an American colony. He thought he was invading a foreign country. He wasn't unusual. Most Americans carried what Daniel Immerwahr calls the logo map: the lower forty-eight, with Alaska and Hawaii tucked into boxes in the corner. It leaves out the Philippines, Guam, Puerto Rico, and the millions of people who lived under the American flag.

The gap shows up on the most famous day of the war. Japan struck the Philippines the same day as Pearl Harbor, and, as Immerwahr shows, that attack was far more devastating. Roosevelt chose which one to emphasize.

The surprise comes after 1945. The United States ended the war with a large empire, half the world's manufacturing, and the atomic bomb. Every precedent said it would take more territory. It gave territory up instead. Immerwahr's explanation is technology. Synthetic rubber meant you didn't need to own rubber plantations. Bases, technical standards, and the English language could project power without holding land.

It reads as more power-hungry than evil, with real restraint at the moment it mattered most. But the Congo and rare-earth minerals still require territory. Technology made some land less valuable, not all of it.

After 1945, power stopped needing a map. Whoever sets the standards and holds the bases runs things now, which is exactly why it's so easy to miss.

In France, workers sometimes strike without walking out. They do exactly what their job descriptions say, nothing more, and everything grinds to a halt. Following the rules perfectly proves the rules don't work.

James Scott builds a whole theory from that insight. Why do grand government plans fail so often? Because the planners imagined themselves far smarter than they were, and imagined their subjects far more stupid and incompetent than they were.

He names four conditions for disaster. Legibility gives a state the capacity, simplifying society so it can be counted. A high-modernist ideology gives it the desire to redesign society from scratch. An authoritarian state gives it the willingness to use force. And a prostrate civil society means nobody can stop it. When all four line up, you get Soviet collectivization, or Tanzania's forced villages.

The word to take away is mētis, the practical knowledge of the people who do the work. The farmer who knows his own soil knows more than the bureaucrat in the capital. Scott's example is Brasília, designed with every good idea in it and, by his account, unpleasant to live in. But this isn't a clean win for the farmer. Some things really are better planned and shared, from public health to Social Security, and Scott says so himself.

Authoritarian states end up in a strange bind. North Korea is almost a hostage situation running in both directions: the government holds the people hostage, and it's so afraid of them that they hold it hostage too.

The state runs what it can see. Everything else, the soil, the side streets, the workarounds that make a plan function, belongs to the people who know it firsthand, and democracy only works if their knowledge is trusted.

At 8:15 on the morning of August 6, 1945, most people in Hiroshima were doing something ordinary. Minutes later, with no warning, they were walking kilometers on catastrophic injuries.

M. G. Sheftall gathered their stories decades later, from survivors who in some cases hadn't spoken about that morning until the 1970s or 80s. Some details are strategic. Sheftall notes that Hiroshima had been spared the firebombing so the bomb's effects could be measured cleanly, and being spared gave people a false sense of safety. Some are cultural, like the raised threshold at the entrance to Japanese homes, a line between order and chaos that people kept even when everything else had fallen apart.

Then there is memory. As Sheftall points out, Hiroshima's memorials almost completely leave out the Akatsuki Command, the military logistics unit that ran much of the city's war effort. After the war, blame settled on a small cabal of officers, which let ordinary Japanese, the emperor, and the Americans all step back from it.

There's no clean verdict here. Japan's own record, from Nanking to the Bataan Death March, doesn't cancel what happened to Hiroshima's civilians, and the reverse doesn't hold either.

A memorial is a verdict too. In Hiroshima, the names left off the stone say as much as the names carved into it.

In October 1943, a company director who had never sailed a boat stood on the Danish coast with his family. The fisherman they had paid to take them to Sweden had backed out, and a German deportation order was already moving. So he asked whether he could pilot the boat himself.

Later in the escape, he set aside cash to repay a family whose home had been damaged in the flight. That detail is the whole rescue in miniature: ordinary people with no plan and no resistance network, doing the next decent thing in front of them.

Almost all of Denmark's Jews reached Sweden that month. Almost nowhere else in occupied Europe came close. Bo Lidegaard's explanation starts years earlier. He argues that Denmark's unpopular cooperation with Hitler was a calculated bet to keep its independence, and the country never accepted the first small steps that elsewhere led to the rest: stars, then property, then internment. The moment to stand up is before the momentum sets in.

The darker lesson comes from the other side. The historian Peter Longerich argues that Nazi leadership deliberately spread a diffuse sense of collective guilt across the German population, so that responsibility could never be pinned to a single decision. When everyone is a little guilty, no one is responsible.

That is one answer to who is responsible when a system does the killing. The system is built so that no one is. Which means responsibility has to be taken, by individuals, early, the way a man who had never sailed offered to take the boat himself.

At Borodino, Napoleon watched from the heights and sent orders to the front. By the time a messenger arrived, the situation had usually changed, and most of the orders were never carried out. Tolstoy uses that scene to argue that great men don't move history. They ride it.

Parts of that case land. Events are multicausal, and once millions of men are moving east and west, it's hard to point to one decision. But one person can matter. In 1983 Stanislav Petrov, a Soviet officer fairly low in the hierarchy, saw what looked like an American missile launch, reported it as a malfunction, and probably prevented a nuclear war. And Napoleon chose the generals who made the decisions at Borodino. Calling him irrelevant is disingenuous.

Tolstoy is most convincing on hierarchy. "The higher they stand in the social hierarchy the less are they free." Power isn't freedom. You trade freedom and accountability for it, and the higher you go, the more you answer to. Kutuzov takes on the job of saving Russia without knowing he'll lose Moscow, and is judged harshly for it.

The second epilogue drops the story for an essay on free will. The objection worth keeping is about inevitability. Things look predestined after the fact, but predicting them ahead of time is one of the things people are worst at, and one of the things we think we're best at.

But the real reason to spend twelve hundred pages with Tolstoy is the people. Pierre drifting from playboy to Freemason to drinking and back to faith. Natasha falling for Anatole the way anyone falls for a shiny object. Petya, whom Tolstoy makes you love in one short chapter before he's shot off his horse. Some people should not be in war. They're too good.

Tolstoy spreads responsibility across millions of people. I'd still hold the people at the top to account, because they chose who carried out their orders.

In a military camp on Rome's northern frontier, in the middle of a long war, the most powerful man in the world was writing notes to himself about staying calm. Marcus Aurelius never meant anyone to read them.

That is why Meditations repeats itself. It's someone thinking out loud for years, not arguing a case. There are what look like contradictions. He says nothing in infinite time will ever recur, and elsewhere that everything repeats. Sit with it and the resolution appears: specific moments never come back, but patterns do.

The practical part is about anxiety. We're one of the only creatures that lives oriented toward the future, bargaining with it and trying to shape it, which is exactly where anxiety comes from. His answer is that your opinion about a thing harms you, not the thing itself. Modern advice about controlling what you can control is the same idea, rediscovered.

Read him against his job, too. An emperor measured against a line of emperors will obsess over legacy. Ideas come from a position, not from nowhere. And much of it will feel familiar, because Stoicism reaches us secondhand, through Epictetus, through Christianity, through a shelf of self-help books.

Marcus rehearsed his values every morning because he expected to be tested every day. Two thousand years later the notes still read as wise, which says something durable about being human, even if it doesn't prove them true.

To a Roman, the cross was the death reserved for slaves and rebels, the most degrading way to die. That a crucified man would become the center of a civilization's moral life would have struck any Roman as absurd.

Tom Holland's argument starts there. List the values you hold most strongly: that the weak have a claim on the strong, that everyone has equal worth regardless of what they can do, that the last should be first. A Roman wouldn't find them merely unfamiliar. He'd find them repellent. Holland's case is that those values come from Christianity, including for the people most uncomfortable with God. Watching that case unfold across two thousand years, from Roman arenas to modern human rights, is a beautiful thing.

Be suspicious of how much you agree, though. If an argument confirms what you already believed, that's a reason for care, not comfort. The real test is whether a belief system with no Abrahamic roots arrives at the same values independently. Islam and Judaism share the trunk, so they don't count. That leaves East Asian traditions like Confucianism, and Holland doesn't test against them. As constructed, the argument may be unfalsifiable.

The sharpest disagreement isn't with Holland at all. It's with Marx, who assumed an early stage of human life when people were equal. Hierarchy might be as old as the animal kingdom, and Christian Europe believed everyone was made in God's image while keeping rigid hierarchy and slavery for well over a thousand years.

A common view treats religion as a bootloader for society, a way to unite people around shared values instead of birthplace or ethnicity. The stronger possibility is that Christianity persists because it maps reality in a way that actually lets a society scale.

So where do our values come from? Holland's answer is Christianity, and it's persuasive. The honest caveat is that if the beliefs fade, we may be living off an inheritance, and it's worth knowing where it came from.