August besieged California with a heat unseen in generations. A surge in air conditioning broke the state’s electrical grid, leaving a population already ravaged by the coronavirus to work remotely by the dim light of their cellphones. By midmonth, the state had recorded possibly the hottest temperature ever measured on earth — 130 degrees in Death Valley — and an otherworldly storm of lightning had cracked open the sky. From Santa Cruz to Lake Tahoe, thousands of bolts of electricity exploded down onto withered grasslands and forests, some of them already hollowed out by climate-driven infestations of beetles and kiln-dried by the worst five-year drought on record. Soon, California was on fire.

Over the next two weeks, 900 blazes incinerated six times as much land as all the state’s 2019 wildfires combined, forcing 100,000 people from their homes. Three of the largest fires in history burned simultaneously in a ring around the San Francisco Bay Area. Another fire burned just 12 miles from my home in Marin County. I watched as towering plumes of smoke billowed from distant hills in all directions and air tankers crisscrossed the skies. Like many Californians, I spent those weeks worrying about what might happen next, wondering how long it would be before an inferno of 60-foot flames swept up the steep, grassy hillside on its way toward my own house, rehearsing in my mind what my family would do to escape.

But I also had a longer-term question, about what would happen once this unprecedented fire season ended. Was it finally time to leave for good?

I had an unusual perspective on the matter. For two years, I have been studying how climate change will influence global migration. My sense was that of all the devastating consequences of a warming planet — changing landscapes, pandemics, mass extinctions — the potential movement of hundreds of millions of climate refugees across the planet stands to be among the most important. I traveled across four countries to witness how rising temperatures were driving climate refugees away from some of the poorest and hottest parts of the world. I had also helped create an enormous computer simulation to analyze how global demographics might shift, and now I was working on a data-mapping project about migration here in the United States.

So it was with some sense of recognition that I faced the fires these last few weeks. In recent years, summer has brought a season of fear to California, with ever-worsening wildfires closing in. But this year felt different. The hopelessness of the pattern was now clear, and the pandemic had already uprooted so many Americans. Relocation no longer seemed like such a distant prospect. Like the subjects of my reporting, climate change had found me, its indiscriminate forces erasing all semblance of normalcy. Suddenly I had to ask myself the very question I’d been asking others: Was it time to move?

I am far from the only American facing such questions. This summer has seen more fires, more heat, more storms — all of it making life increasingly untenable in larger areas of the nation. Already, droughts regularly threaten food crops across the West, while destructive floods inundate towns and fields from the Dakotas to Maryland, collapsing dams in Michigan and raising the shorelines of the Great Lakes. Rising seas and increasingly violent hurricanes are making thousands of miles of American shoreline nearly uninhabitable. As California burned, Hurricane Laura pounded the Louisiana coast with 150-mile-an-hour winds, killing at least 25 people; it was the 12th named storm to form by that point in 2020, another record. Phoenix, meanwhile, endured 53 days of 110-degree heat — 20 more days than the previous record.

For years, Americans have avoided confronting these changes in their own backyards. The decisions we make about where to live are distorted not just by politics that play down climate risks, but also by expensive subsidies and incentives aimed at defying nature. In much of the developing world, vulnerable people will attempt to flee the emerging perils of global warming, seeking cooler temperatures, more fresh water and safety. But here in the United States, people have largely gravitated toward environmental danger, building along coastlines from New Jersey to Florida and settling across the cloudless deserts of the Southwest.

I wanted to know if this was beginning to change. Might Americans finally be waking up to how climate is about to transform their lives? And if so — if a great domestic relocation might be in the offing — was it possible to project where we might go? To answer these questions, I interviewed more than four dozen experts: economists and demographers, climate scientists and insurance executives, architects and urban planners, and ProPublica mapped out the danger zones that will close in on Americans over the next 30 years. The maps for the first time combined exclusive climate data from the Rhodium Group, an independent data-analytics firm; wildfire projections modeled by United States Forest Service researchers and others; and data about America’s shifting climate niches, an evolution of work first published by the Proceedings of the National Academy of Sciences last spring. (A detailed analysis of the maps is available here.)

What I found was a nation on the cusp of a great transformation. Across the United States, some 162 million people — nearly 1 in 2 — will most likely experience a decline in the quality of their environment, namely more heat and less water. For 93 million of them, the changes could be particularly severe, and by 2070, our analysis suggests, if carbon emissions rise at extreme levels, at least 4 million Americans could find themselves living at the fringe, in places decidedly outside the ideal niche for human life. The cost of resisting the new climate reality is mounting. Florida officials have already acknowledged that defending some roadways against the sea will be unaffordable. And the nation’s federal flood-insurance program is for the first time requiring that some of its payouts be used to retreat from climate threats across the country. It will soon prove too expensive to maintain the status quo.

Then what? One influential 2018 study, published in the Journal of the Association of Environmental and Resource Economists, suggests that 1 in 12 Americans in the Southern half of the country will move toward California, the Mountain West or the Northwest over the next 45 years because of climate influences alone. Such a shift in population is likely to increase poverty and widen the gulf between the rich and the poor. It will accelerate rapid, perhaps chaotic, urbanization of cities ill-equipped for the burden, testing their capacity to provide basic services and amplifying existing inequities. It will eat away at prosperity, dealing repeated economic blows to coastal, rural and Southern regions, which could in turn push entire communities to the brink of collapse. This process has already begun in rural Louisiana and coastal Georgia, where low-income and Black and Indigenous communities face environmental change on top of poor health and extreme poverty. Mobility itself, global-migration experts point out, is often a reflection of relative wealth, and as some move, many others will be left behind. Those who stay risk becoming trapped as the land and the society around them ceases to offer any more support.

There are signs that the message is breaking through. Half of Americans now rank climate as a top political priority, up from roughly one-third in 2016, and 3 out of 4 now describe climate change as either “a crisis” or “a major problem.” This year, Democratic caucusgoers in Iowa, where tens of thousands of acres of farmland flooded in 2019, ranked climate second only to health care as an issue. A poll by researchers at Yale and George Mason universities found that even Republicans’ views are shifting: 1 in 3 now thinks climate change should be declared a national emergency.

Policymakers, having left America unprepared for what’s next, now face brutal choices about which communities to save — often at exorbitant costs — and which to sacrifice. Their decisions will almost inevitably make the nation more divided, with those worst off relegated to a nightmare future in which they are left to fend for themselves. Nor will these disruptions wait for the worst environmental changes to occur. The wave begins when individual perception of risk starts to shift, when the environmental threat reaches past the least fortunate and rattles the physical and financial security of broader, wealthier parts of the population. It begins when even places like California’s suburbs are no longer safe.

It has already begun.

---

Let’s start with some basics. Across the country, it’s going to get hot. Buffalo, New York, may feel in a few decades like Tempe, Arizona, does today, and Tempe itself will sustain 100-degree average summer temperatures by the end of the century. Extreme humidity from New Orleans to northern Wisconsin will make summers increasingly unbearable, turning otherwise seemingly survivable heat waves into debilitating health threats. Fresh water will also be in short supply, not only in the West but also in places like Florida, Georgia and Alabama, where droughts now regularly wither cotton fields. By 2040, according to federal government projections, extreme water shortages will be nearly ubiquitous west of Missouri. The Memphis Sands Aquifer, a crucial water supply for Mississippi, Tennessee, Arkansas and Louisiana, is already overdrawn by hundreds of millions of gallons a day. Much of the Ogallala Aquifer — which supplies nearly a third of the nation’s irrigation groundwater — could be gone by the end of the century.

It can be difficult to see the challenges clearly because so many factors are in play. At least 28 million Americans are likely to face megafires like the ones we are now seeing in California, in places like Texas and Florida and Georgia. At the same time, 100 million Americans — largely in the Mississippi River Basin from Louisiana to Wisconsin — will increasingly face humidity so extreme that working outside or playing school sports could cause heatstroke. Crop yields will be decimated from Texas to Alabama and all the way north through Oklahoma and Kansas and into Nebraska.

The challenges are so widespread and so interrelated that Americans seeking to flee one could well run into another. I live on a hilltop, 400 feet above sea level, and my home will never be touched by rising waters. But by the end of this century, if the more extreme projections of 8 to 10 feet of sea-level rise come to fruition, the shoreline of San Francisco Bay will move 3 miles closer to my house, as it subsumes some 166 square miles of land, including a high school, a new county hospital and the store where I buy groceries. The freeway to San Francisco will need to be raised, and to the east, a new bridge will be required to connect the community of Point Richmond to the city of Berkeley. The Latino, Asian and Black communities who live in the most-vulnerable low-lying districts will be displaced first, but research from Mathew Hauer, a sociologist at Florida State University who published some of the first modeling of American climate migration in the journal Nature Climate Change in 2017, suggests that the toll will eventually be far more widespread: Nearly 1 in 3 people here in Marin County will leave, part of the roughly 700,000 who his models suggest may abandon the broader Bay Area as a result of sea-level rise alone.

From Maine to North Carolina to Texas, rising sea levels are not just chewing up shorelines but also raising rivers and swamping the subterranean infrastructure of coastal communities, making a stable life there all but impossible. Coastal high points will be cut off from roadways, amenities and escape routes, and even far inland, saltwater will seep into underground drinking-water supplies. Eight of the nation’s 20 largest metropolitan areas — Miami, New York and Boston among them — will be profoundly altered, indirectly affecting some 50 million people. Imagine large concrete walls separating Fort Lauderdale, Florida, condominiums from a beachless waterfront, or dozens of new bridges connecting the islands of Philadelphia. Not every city can spend $100 billion on a sea wall, as New York most likely will. Barrier islands? Rural areas along the coast without a strong tax base? They are likely, in the long term, unsalvageable.

In all, Hauer projects that 13 million Americans will be forced to move away from submerged coastlines. Add to that the people contending with wildfires and other risks, and the number of Americans who might move — though difficult to predict precisely — could easily be tens of millions larger. Even 13 million climate migrants, though, would rank as the largest migration in North American history. The Great Migration — of 6 million Black Americans out of the South from 1916 to 1970 — transformed almost everything we know about America, from the fate of its labor movement to the shape of its cities to the sound of its music. What would it look like when twice that many people moved? What might change?

---

Americans have been conditioned not to respond to geographical climate threats as people in the rest of the world do. It is natural that rural Guatemalans or subsistence farmers in Kenya, facing drought or scorching heat, would seek out someplace more stable and resilient. Even a subtle environmental change — a dry well, say — can mean life or death, and without money to address the problem, migration is often simply a question of survival.

By comparison, Americans are richer, often much richer, and more insulated from the shocks of climate change. They are distanced from the food and water sources they depend on, and they are part of a culture that sees every problem as capable of being solved by money. So even as the average flow of the Colorado River — the water supply for 40 million Western Americans and the backbone of the nation’s vegetable and cattle farming — has declined for most of the last 33 years, the population of Nevada has doubled. At the same time, more than 1.5 million people have moved to the Phoenix metro area, despite its dependence on that same river (and the fact that temperatures there now regularly hit 115 degrees). Since Hurricane Andrew devastated Florida in 1992 — and even as that state has become a global example of the threat of sea-level rise — more than 5 million people have moved to Florida’s shorelines, driving a historic boom in building and real estate.

Similar patterns are evident across the country. Census data shows us how Americans move: toward heat, toward coastlines, toward drought, regardless of evidence of increasing storms and flooding and other disasters.

The sense that money and technology can overcome nature has emboldened Americans. Where money and technology fail, though, it inevitably falls to government policies — and government subsidies — to pick up the slack. Thanks to federally subsidized canals, for example, water in part of the Desert Southwest costs less than it does in Philadelphia. The federal National Flood Insurance Program has paid to rebuild houses that have flooded six times over in the same spot. And federal agriculture aid withholds subsidies from farmers who switch to drought-resistant crops, while paying growers to replant the same ones that failed. Farmers, seed manufacturers, real estate developers and a few homeowners benefit, at least momentarily, but the gap between what the climate can destroy and what money can replace is growing.

Perhaps no market force has proved more influential — and more misguided — than the nation’s property-insurance system. From state to state, readily available and affordable policies have made it attractive to buy or replace homes even where they are at high risk of disasters, systematically obscuring the reality of the climate threat and fooling many Americans into thinking that their decisions are safer than they actually are. Part of the problem is that most policies look only 12 months into the future, ignoring long-term trends even as insurance availability influences development and drives people’s long-term decision-making.

Even where insurers have tried to withdraw policies or raise rates to reduce climate-related liabilities, state regulators have forced them to provide affordable coverage anyway, simply subsidizing the cost of underwriting such a risky policy or, in some cases, offering it themselves. The regulations — called Fair Access to Insurance Requirements — are justified by developers and local politicians alike as economic lifeboats “of last resort” in regions where climate change threatens to interrupt economic growth. While they do protect some entrenched and vulnerable communities, the laws also satisfy the demand of wealthier homeowners who still want to be able to buy insurance.

At least 30 states, including Louisiana, Massachusetts, North Carolina and Texas, have developed so-called FAIR plans, and today they serve as a market backstop in the places facing the highest risks of climate-driven disasters, including coastal flooding, hurricanes and wildfires.

In an era of climate change, though, such policies amount to a sort of shell game, meant to keep growth going even when other obvious signs and scientific research suggest that it should stop.

That’s what happened in Florida. Hurricane Andrew reduced parts of cities to landfill and cost insurers nearly $16 billion in payouts. Many insurance companies, recognizing the likelihood that it would happen again, declined to renew policies and left the state. So the Florida Legislature created a state-run company to insure properties itself, preventing both an exodus and an economic collapse by essentially pretending that the climate vulnerabilities didn’t exist.

As a result, Florida’s taxpayers by 2012 had assumed liabilities worth some $511 billion — more than seven times the state’s total budget — as the value of coastal property topped $2.8 trillion. Another direct hurricane risked bankrupting the state. Florida, concerned that it had taken on too much risk, has since scaled back its self-insurance plan. But the development that resulted is still in place.

---

On a sweltering afternoon last October, with the skies above me full of wildfire smoke, I called Jesse Keenan, an urban-planning and climate-change specialist then at Harvard’s Graduate School of Design, who advises the federal Commodity Futures Trading Commission on market hazards from climate change. Keenan, who is now an associate professor of real estate at Tulane University’s School of Architecture, had been in the news last year for projecting where people might move to — suggesting that Duluth, Minnesota, for instance, should brace for a coming real estate boom as climate migrants move north. But like other scientists I’d spoken with, Keenan had been reluctant to draw conclusions about where these migrants would be driven from.

Last fall, though, as the previous round of fires ravaged California, his phone began to ring, with private-equity investors and bankers all looking for his read on the state’s future. Their interest suggested a growing investor-grade nervousness about swiftly mounting environmental risk in the hottest real estate markets in the country. It’s an early sign, he told me, that the momentum is about to switch directions. “And once this flips,” he added, “it’s likely to flip very quickly.”

In fact, the correction — a newfound respect for the destructive power of nature, coupled with a sudden disavowal of Americans’ appetite for reckless development — had begun two years earlier, when a frightening surge in disasters offered a jolting preview of how the climate crisis was changing the rules.

On Oct. 9, 2017, a wildfire blazed through the suburban blue-collar neighborhood of Coffey Park in Santa Rosa, California, virtually in my own backyard. I awoke to learn that more than 1,800 buildings were reduced to ashes, less than 35 miles from where I slept. Inchlong cinders had piled on my windowsills like falling snow.

The Tubbs Fire, as it was called, shouldn’t have been possible. Coffey Park is surrounded not by vegetation but by concrete and malls and freeways. So insurers had rated it as “basically zero risk,” according to Kevin Van Leer, then a risk modeler from the global insurance liability firm Risk Management Solutions. (He now does similar work for Cape Analytics.) But Van Leer, who had spent seven years picking through the debris left by disasters to understand how insurers could anticipate — and price — the risk of their happening again, had begun to see other “impossible” fires. After a 2016 fire tornado ripped through northern Canada and a firestorm consumed Gatlinburg, Tennessee, he said, “alarm bells started going off” for the insurance industry.

What Van Leer saw when he walked through Coffey Park a week after the Tubbs Fire changed the way he would model and project fire risk forever. Typically, fire would spread along the ground, burning maybe 50% of structures. In Santa Rosa, more than 90% had been leveled. “The destruction was complete,” he told me. Van Leer determined that the fire had jumped through the forest canopy, spawning 70-mile-per-hour winds that kicked a storm of embers into the modest homes of Coffey Park, which burned at an acre a second as homes ignited spontaneously from the radiant heat. It was the kind of thing that might never have been possible if California’s autumn winds weren’t getting fiercer and drier every year, colliding with intensifying, climate-driven heat and ever-expanding development. “It’s hard to forecast something you’ve never seen before,” he said.

For me, the awakening to imminent climate risk came with California’s rolling power blackouts last fall — an effort to preemptively avoid the risk of a live wire sparking a fire — which showed me that all my notional perspective about climate risk and my own life choices were on a collision course. After the first one, all the food in our refrigerator was lost. When power was interrupted six more times in three weeks, we stopped trying to keep it stocked. All around us, small fires burned. Thick smoke produced fits of coughing. Then, as now, I packed an ax and a go-bag in my car, ready to evacuate. As former Gov. Jerry Brown said, it was beginning to feel like the “new abnormal.”

It was no surprise, then, that California’s property insurers — having watched 26 years’ worth of profits dissolve over 24 months — began dropping policies, or that California’s insurance commissioner, trying to slow the slide, placed a moratorium on insurance cancellations for parts of the state in 2020. In February, the Legislature introduced a bill compelling California to, in the words of one consumer advocacy group, “follow the lead of Florida” by mandating that insurance remain available, in this case with a requirement that homeowners first harden their properties against fire. At the same time, participation in California’s FAIR plan for catastrophic fires has grown by at least 180% since 2015, and in Santa Rosa, houses are being rebuilt in the very same wildfire-vulnerable zones that proved so deadly in 2017. Given that a new study projects a 20% increase in extreme-fire-weather days by 2035, such practices suggest a special form of climate negligence.

It’s only a matter of time before homeowners begin to recognize the unsustainability of this approach. Market shock, when driven by the sort of cultural awakening to risk that Keenan observes, can strike a neighborhood like an infectious disease, with fear spreading doubt — and devaluation — from door to door. It happened that way in the foreclosure crisis.

Keenan calls the practice of drawing arbitrary lending boundaries around areas of perceived environmental risk “bluelining,” and indeed many of the neighborhoods that banks are bluelining are the same as the ones that were hit by the racist redlining practice in days past. This summer, climate-data analysts at the First Street Foundation released maps showing that 70% more buildings in the United States were vulnerable to flood risk than previously thought; most of the underestimated risk was in low-income neighborhoods.

Such neighborhoods see little in the way of flood-prevention investment. My Bay Area neighborhood, on the other hand, has benefited from consistent investment in efforts to defend it against the ravages of climate change. That questions of livability had reached me, here, were testament to Keenan’s belief that the bluelining phenomenon will eventually affect large majorities of equity-holding middle-class Americans too, with broad implications for the overall economy, starting in the nation’s largest state.

Under the radar, a new class of dangerous debt — climate-distressed mortgage loans — might already be threatening the financial system. Lending data analyzed by Keenan and his co-author, Jacob Bradt, for a study published in the journal Climatic Change in June shows that small banks are liberally making loans on environmentally threatened homes, but then quickly passing them along to federal mortgage backers. At the same time, they have all but stopped lending money for the higher-end properties worth too much for the government to accept, suggesting that the banks are knowingly passing climate liabilities along to taxpayers as stranded assets.

Once home values begin a one-way plummet, it’s easy for economists to see how entire communities spin out of control. The tax base declines and the school system and civic services falter, creating a negative feedback loop that pushes more people to leave. Rising insurance costs and the perception of risk force credit-rating agencies to downgrade towns, making it more difficult for them to issue bonds and plug the springing financial leaks. Local banks, meanwhile, keep securitizing their mortgage debt, sloughing off their own liabilities.

Keenan, though, had a bigger point: All the structural disincentives that had built Americans’ irrational response to the climate risk were now reaching their logical endpoint. A pandemic-induced economic collapse will only heighten the vulnerabilities and speed the transition, reducing to nothing whatever thin margin of financial protection has kept people in place. Until now, the market mechanisms had essentially socialized the consequences of high-risk development. But as the costs rise — and the insurers quit, and the bankers divest, and the farm subsidies prove too wasteful, and so on — the full weight of responsibility will fall on individual people.

And that’s when the real migration might begin.

As I spoke with Keenan last year, I looked out my own kitchen window onto hillsides of parkland, singed brown by months of dry summer heat. This was precisely the land that my utility, Pacific Gas & Electric, had three times identified as such an imperiled tinderbox that it had to shut off power to avoid fire. It was precisely the kind of wildland-urban interface that all the studies I read blamed for heightening Californians’ exposure to climate risks. I mentioned this on the phone and then asked Keenan, “Should I be selling my house and getting — ”

He cut me off: “Yes.”

---

Americans have dealt with climate disaster before. The Dust Bowl started after the federal government expanded the Homestead Act to offer more land to settlers willing to work the marginal soil of the Great Plains. Millions took up the invitation, replacing hardy prairie grass with thirsty crops like corn, wheat and cotton. Then, entirely predictably, came the drought. From 1929 to 1934, crop yields across Texas, Oklahoma, Kansas and Missouri plunged by 60%, leaving farmers destitute and exposing the now-barren topsoil to dry winds and soaring temperatures. The resulting dust storms, some of them taller than skyscrapers, buried homes whole and blew as far east as Washington. The disaster propelled an exodus of some 2.5 million people, mostly to the West, where newcomers — “Okies” not just from Oklahoma but also Texas, Arkansas and Missouri — unsettled communities and competed for jobs. Colorado tried to seal its border from the climate refugees; in California, they were funneled into squalid shanty towns. Only after the migrants settled and had years to claw back a decent life did some towns bounce back stronger.

The places migrants left behind never fully recovered. Eighty years later, Dust Bowl towns still have slower economic growth and lower per capita income than the rest of the country. Dust Bowl survivors and their children are less likely to go to college and more likely to live in poverty. Climatic change made them poor, and it has kept them poor ever since.

A Dust Bowl event will most likely happen again. The Great Plains states today provide nearly half of the nation’s wheat, sorghum and cattle and much of its corn; the farmers and ranchers there export that food to Africa, South America and Asia. Crop yields, though, will drop sharply with every degree of warming. By 2050, researchers at the University of Chicago and the NASA Goddard Institute for Space Studies found, Dust Bowl-era yields will be the norm, even as demand for scarce water jumps by as much as 20%. Another extreme drought would drive near-total crop losses worse than the Dust Bowl, kneecapping the broader economy. At that point, the authors write, “abandonment is one option.”

Projections are inherently imprecise, but the gradual changes to America’s cropland — plus the steady baking and burning and flooding — suggest that we are already witnessing a slower-forming but much larger replay of the Dust Bowl that will destroy more than just crops. In 2017, Solomon Hsiang, a climate economist at the University of California, Berkeley, led an analysis of the economic impact of climate-driven changes like rising mortality and rising energy costs, finding that the poorest counties in the United States — mostly across the South and the Southwest — will in some extreme cases face damages equal to more than a third of their gross domestic products. The 2018 National Climate Assessment also warns that the U.S. economy over all could contract by 10%.

That kind of loss typically drives people toward cities, and researchers expect that trend to continue after the COVID-19 pandemic ends. In 1950, less than 65% of Americans lived in cities. By 2050, only 10% will live outside them, in part because of climatic change. By 2100, Hauer estimates, Atlanta, Orlando, Houston and Austin could each receive more than a quarter million new residents as a result of sea-level displacement alone, meaning it may be those cities — not the places that empty out — that wind up bearing the brunt of America’s reshuffling. The World Bank warns that fast-moving climate urbanization leads to rising unemployment, competition for services and deepening poverty.

So what will happen to Atlanta — a metro area of 5.8 million people that may lose its water supply to drought and that our data also shows will face an increase in heat-driven wildfires? Hauer estimates that hundreds of thousands of climate refugees will move into the city by 2100, swelling its population and stressing its infrastructure. Atlanta — where poor transportation and water systems contributed to the state’s C+ infrastructure grade last year — already suffers greater income inequality than any other large American city, making it a virtual tinderbox for social conflict. One in 10 households earns less than $10,000 a year, and rings of extreme poverty are growing on its outskirts even as the city center grows wealthier.

Atlanta has started bolstering its defenses against climate change, but in some cases this has only exacerbated divisions. When the city converted an old Westside rock quarry into a reservoir, part of a larger greenbelt to expand parkland, clean the air and protect against drought, the project also fueled rapid upscale growth, driving the poorest Black communities further into impoverished suburbs. That Atlanta hasn’t “fully grappled with” such challenges now, said Na’Taki Osborne Jelks, chair of the West Atlanta Watershed Alliance, means that with more people and higher temperatures, “the city might be pushed to what’s manageable.”

So might Philadelphia, Chicago, Washington, Boston and other cities with long-neglected systems suddenly pressed to expand under increasingly adverse conditions.

---

Once you accept that climate change is fast making large parts of the United States nearly uninhabitable, the future looks like this: With time, the bottom half of the country grows inhospitable, dangerous and hot. Something like a tenth of the people who live in the South and the Southwest — from South Carolina to Alabama to Texas to Southern California — decide to move north in search of a better economy and a more temperate environment. Those who stay behind are disproportionately poor and elderly.

In these places, heat alone will cause as many as 80 additional deaths per 100,000 people — the nation’s opioid crisis, by comparison, produces 15 additional deaths per 100,000. The most affected people, meanwhile, will pay 20% more for energy, and their crops will yield half as much food or in some cases virtually none at all. That collective burden will drag down regional incomes by roughly 10%, amounting to one of the largest transfers of wealth in American history, as people who live farther north will benefit from that change and see their fortunes rise.

The millions of people moving north will mostly head to the cities of the Northeast and Northwest, which will see their populations grow by roughly 10%, according to one model. Once-chilly places like Minnesota and Michigan and Vermont will become more temperate, verdant and inviting. Vast regions will prosper; just as Hsiang’s research forecast that Southern counties could see a tenth of their economy dry up, he projects that others as far as North Dakota and Minnesota will enjoy a corresponding expansion. Cities like Detroit; Rochester, New York; Buffalo and Milwaukee will see a renaissance, with their excess capacity in infrastructure, water supplies and highways once again put to good use. One day, it’s possible that a high-speed rail line could race across the Dakotas, through Idaho’s up-and-coming wine country and the country’s new breadbasket along the Canadian border, to the megalopolis of Seattle, which by then has nearly merged with Vancouver to its north.

Sitting in my own backyard one afternoon this summer, my wife and I talked through the implications of this looming American future. The facts were clear and increasingly foreboding. Yet there were so many intangibles — a love of nature, the busy pace of life, the high cost of moving — that conspired to keep us from leaving. Nobody wants to migrate away from home, even when an inexorable danger is inching ever closer. They do it when there is no longer any other choice.