Wie het water deert, die het water keert. (Whom the water hurts, he the water stops.)
—Old Dutch proverb
IN TELLING THE STORY OF THE SEARCH FOR THE CURE FOR catastrophe, we have focused so far on the professionals: the scientists, the builders, the insurers, and the risk modelers. Now the curtain draws back—the scene is about to expand as we explore the strategies that societies have learned to manage their disasters. Some of those strategies have been broadly successful, while others have turned out to be misguided or impractical. We will start with some successes.
One country stands out for the way it made action on disasters core to the functioning of the state. That country is the Netherlands.
Four hundred years ago, the Dutch had mastered floods. They had come to understand that high river levels have to be guided to the sea, while sea floods need to be blocked. They had learned how to construct strong, wave-resistant embankments and where to mount secondary lines of backstop defenses. They had grown an operational system that could call out every villager to fight a dike breach, while mobilizing the ruling prince to provide the money and manpower to defend a threatened coastline. With the will, the funding, the leadership, and the communal ownership, a nation could beat disasters. The defeat of the storm surges and the elevated status given to national water management defined the Dutch nation and its entrepreneurial and consensual government. “Polder democracy” still flourishes in Holland today. In the words of the Dutch historian Herman Pleij: “The Netherlands owes its existence to the democracy of dry feet. We need each other literally in order not to drown.”1
A thousand years ago, at the end of its journey from the Alps to the sea, the River Rhine split into three channels as it flowed through the “great bog of Europe” (as it was later derided), a vast delta of swamplands, thickets, and raised peat mires extending 50 miles (80 kilometers) inland from the North Sea coast. Apart from the coastal dunes, the land that today forms the Netherlands was completely unrecognizable then: there were no canals, windmills, or dikes, no settlement at Amsterdam, no inland Zuiderzee. The tribes who lived around the great bog huddled in their huts on dwelling mounds (terpen) raised above the tidal marshes and sustained themselves through fishing and trapping.2
Starting around AD 1000, the chief noble of the region, styling himself the Count of Holland, encouraged free peasants to create new farmland.3 First they cut a deep ditch, fed by regularly spaced side ditches that each bounded a rented land parcel. The ditches would “bleed” the bogs, and in a generation, as the peat oxidized and the ground subsided, there would be fertile soil for growing crops or fodder for livestock. Around the same time, settlers living closer to the coast raised embankments to turn tidal salt marshes into grazing land.4 Tens of thousands of people came to live on the drained marshlands, in farms, monasteries, and villages of single-story thatched cob houses, close to their land and their animals. Within two centuries, this fertile plain had become one of the most densely populated areas of Europe.
As the ground sank, the ditches had to be dug deeper. By the end of the twelfth century, this work required a dedicated team—the heemraadschappen (“water guardians”), who were elected from each village to maintain the drainage on behalf of the village.5 They faced some daunting challenges. Runoff from rain had to be guided along the ditches to the rivers, but river floods threatened to flow back over the land. Worse, over large areas the drained bogs were sinking beneath the level of the highest tides.
In the All Saints’ Day Flood of 1170, the sea surged through the Creiler forest to link up with the large freshwater Almere Lake and then would not recede; instead, the flooded area expanded over the months and years as the waves eroded the shallow layers of peat, forming a great inland sea—the Zuiderzee.6 The Dutch settlers were forced to retreat. For the first time ships could now sail into the interior of the country, and at the end of the thirteenth century, where the River Amstel creek flowed into the new tidal sea, they built a dam and sluice to regulate the level of water in the river. The village where goods were transferred from seagoing sailing ships to inland barges was called Amster-dam.7
Each abbey, farm, and village was responsible for maintaining and repairing the dike that protected them on the principle cited at the beginning of this chapter: whom the water hurts, he the water stops.8 Yet these embankments, being only high enough to protect against ordinary high water, were prone to being undermined or overtopped.
This was the beginning of a 400-year war. Decades of slow reclamation would be reversed by catastrophes when thousands of acres were lost in a single night. Through the thirteenth and fourteenth centuries, it is estimated, more than 200,000 people were drowned by storm surge floods.9
By 1300, the village water guardians were networked into the regional administration and by 1350 had been given the power to raise taxes.10 The system was effective because it was ruthlessly managed. The supervisory “dike count” came to inspect every flood defense at the end of the winter and itemized the necessary repairs; returning in early autumn to check whether they had been completed, he would impose large fines and confiscate land for noncompliance.11 The dike count might be the only source of funds for making the repairs, applying interest rates of 100 percent or more and requisitioning property when debts or fines went unpaid.
The nearest the whole system came to breakdown was at the beginning of the fifteenth century, especially at the time of the great St. Elisabeth’s Day Flood of 1421.12 In a single night, a wide new arm of the sea was carved out of the Scheldt estuary. Inland, between the Waal and the Maas Rivers, an area the size of a small county (500 square kilometers) was overwhelmed and lost; rich and densely populated lands, including forty-eight villages, were flooded so deeply that only the church spires were visible.
After the flood, defeatism took root: people were unwilling to repair dikes and disputed the response. Fortunately, the country’s foreign rulers rose to the self-interested task of protecting their inheritance. Along the coast, the sea-fronting dikes were built higher and stronger.
Since the fourteenth century, people had dug peat for fuel on an industrial scale. The shallow pits became lakes, which expanded with the waves and threatened to undermine the country from within. In the early fifteenth century, a technological solution was found to this insidious problem: windmill-powered pumps that could finally reclaim land below the low tide level.13
In a 1565 painting by Pieter Bruegel the Elder, The Gloomy Day (one of the five surviving “season” pictures), the right half captures a spit of delta land embanked with sea defenses, on which people have built their houses. A winter storm is raging, ships are wrecked along the dike, and the land has partly flooded. Across the estuary there is a typical village, a cluster of thatched buildings, some half-timbered, with pink plaster walls.
The scene was painted five years before these coastal dikes were widely breached and overwhelmed in the 1570 flood, when water levels rose higher than at any time until the present day: 13.5 feet (4 meters) above sea level in places and 11.5 feet (3.5 meters) above sea level in Antwerp.14 The flood inundated the streets of Dordrecht and Rotterdam, killing 3,000 people in the Zeeland islands, while tens of thousands of people drowned inland. Not a single island remained unflooded. Everywhere houses were washed away in the dunes, and whole streets disappeared in Scheveningen. Yet, while many dikes failed, only a few small Zeeland islands were permanently lost.
The tide of the battle with the floods was finally turned in the second half of the sixteenth century. This communal endeavor defined the new nation: “The making of new land belongs to God alone . . . for He gives to some people the wit and strength to do it.” Or more brazenly: “God made the World, but the Dutch made Holland.”15 At the end of the sixteenth century, the dike builder Andries Vierlingh described the responsibility of everyone in the village to seal a breaching dike: “I sent the drummer around to summon every person to work. . . . Those women who had no barrows carried clay in their aprons.”16
Investment moved beyond repairing the damage caused by the last breach. Dikes were faced with timber imported from Norway, and some river dikes were lined with stone brought down the River Meuse by barge. Born in 1575, Jan Adriaanszoon, the lead engineer in the “conquest” of the Beemster Lake (an inland sea north of Amsterdam), changed his name as an act of self-promotion to Leeghwater (Low-water) and was as famous in Holland as the most victorious admiral.17 Forty-three water-raising windmills created 17,500 acres of new fertile farmland, enriching the scheme’s funders with annual rents worth 17 percent of their original investment. Amsterdam’s population soared fivefold between 1578 and 1648, from 31,000 to 150,000, and was sustained by all the food grown on the newly protected farmland. A scheduled barge departed from Amsterdam to Haarlem every hour, using a network of canals that interlinked all the principal cities of the United Provinces.
By 1688, the Dutch had the highest per capita income of any northern European state.18 The prosperity of the late-seventeenth-century Dutch Golden Age could not have been achieved without this victory over nature.
The story of Dutch flood control is the archetypal “cure for catastrophe” parable. The problem was never simply solved once and for all, but required constant vigilance in every new generation. The timber protection on dikes introduced in the seventeenth century was later found to concentrate the energy of breaking waves, eroding the supporting earth. Then, in the eighteenth century, an expensive switch to imported stone facing was required when dike timbers became riddled by the shipworm Teredo navalis.
In the nineteenth century, the regional water boards were combined into the national Rijkswaterstaat.19 It took the 1953 floods, when 1,800 died, for the government to decree that the level of risk of flooding from the sea—the level of intolerable risk—should be set by statute to below one in 10,000 per year.20 Today one-quarter of the Netherlands is below mean sea level, and another two-fifths is at or just above this level, protected by the great coastal wall of sand dunes and the defensive dike rings.
WHAT OTHER COUNTRY HAS REACHED THAT THRESHOLD OF INTOLERABLE risk and made it a societal mission to drive down the forces of destruction? One might not think the rumpled mountain landscape of Japan has anything in common with the canal-seamed plains of the Netherlands, but half the population and three-quarters of Japan’s assets are situated in coastal and river floodplains, which make up only 14 percent of the land area. Japan is Holland with mountains in the middle. Typhoon-driven storm surges are at least as big as those in Holland, while in the summer in Japan, especially over the mountains, the rain can be torrential.
The first River Law, passed in 1896, required levees to be constructed along the banks of the lower channels of all the major rivers.21 Meandering channels were straightened to help the floods flow straight out to sea. The work was completed by 1930, yet within a decade an average of 200,000 properties were being flooded each year.
Forty days after the atomic bomb was dropped on the city of Hiroshima, Typhoon Makurazaki passed over the city—the second-most-intense storm to hit Japan in modern times. The deluge drowned 2,000 survivors of the bomb, who were hiding in all that remained of their buildings—the basements.22 In September 1947, the rainfall from Typhoon Kathleen in the Kantō Plain caused the River Tone embankment to burst, flooding the whole of eastern Tokyo for five days and drowning more than 1,000 people. Through the 1950s, 3 million properties were flooded.
The decade culminated in Japan’s greatest flood disaster of all time. On September 26, 1959, a strong typhoon pushed an 11.5-foot (3.5-meter) surge tide into Ise Bay, overwhelming the defenses on the west side of the city of Nagoya and drowning more than 5,000.23 Tens of thousands of tree trunks that had been stored for builders became wave-driven battering rams. One-fifth of the 250,000 houses in Nagoya were ruined.24 Land subsidence from groundwater extraction left a lake one-third the size of Inner Tokyo Bay, with water up to 10 feet (3 meters) deep. It took three months to pump the water out.
The Ise Bay disaster provoked the Japanese to declare their flood risk “intolerable.” The Disaster Countermeasures Basic Act, passed in 1961, created the Central Disaster Prevention Council, chaired by the prime minister.25 In 1961 almost 8 percent of the national budget (1.5 percent of the gross domestic product) was allocated to disaster prevention. By the late 1980s, against a backdrop of rapid economic growth, a disaster management budget of 4 trillion yen (around $40 billion) still consumed 4.5 percent of the national budget (0.5 percent of GDP).
After three decades of investment in concrete flood walls and retention basins, the annual number of flooded properties fell below 40,000. By the 1990s, flood casualties were only a small fraction of those incurred in the 1950s.26
Yet, as with all flood walls, no protection is absolute, and on the rare occasion when they do fail—as in New Orleans—the higher the wall the deeper the flood.27 In September 2000, after 24 inches (600 millimeters) of rainfall, the Shonai River, confined between its tall flood defenses, flowed 16 feet (5 meters) above the surrounding Nagoya suburbs. When a single section of the wall collapsed, 70,000 houses and businesses and 100,000 cars were inundated.28
THE CURE FOR EARTHQUAKES PROVED FAR HARDER TO FIND THAN the cure for floods. A strong city wall can keep wolves, marauders, and storm tides at bay, but no wall keeps out earthquakes. For centuries, defensive strategies were based on the latest earthquake theory, applied during reconstruction. Whatever these interventions accomplished in terms of economic renewal was eventually diminished by how little they achieved in the way of reducing risk.
If the original site of the city was doomed to disaster, why not move to a new site, innocent of bad associations? From the end of the seventeenth century, in both the Old and New Worlds, relocation became a standard Spanish colonial strategy as a remedy for a devastated city. While such moves demonstrated purposeful leadership, they often proved deeply unpopular.
There were rich opportunities to apply this geomancy after the January 1693 destruction of forty-nine cities across southeast Sicily. Among these was Noto, situated on the edge of the mountains.29 The original Bronze Age site had been selected for its defensive qualities: the Monte Alveria plateau was incised like a maple leaf by ravines. Home to 10,000 people, the city was sustained by fifty-six churches and nineteen convents and monasteries. Then came the earthquake of January 9, 1693, which was “so horrible and ghastly that the soil undulated like the waves of a stormy sea and the mountains danced as if drunk, and the city collapsed in one miserable moment killing more than a thousand people.”30 This was the biggest earthquake in Italy’s history.31 The destruction was so widespread across Sicily that no one came to help.
Sicily was reluctantly ruled from Spain. Over the previous half-century, the province had faced uprisings and economic decline. The Spanish viceroy in Sicily appointed the duke of Camastra to the role of disaster manager. Camastra badgered the people of Noto to endorse relocation to a hilltop halfway to the coast, close to the coastal marshes. In the first summer after they moved, 3,000 died in a malaria epidemic. The survivors clamored to return to Old Noto.32 Eventually people were coerced to stay, after watching the dressed stone from their old houses being transported to the new site in thousands of mule trains.
The buildings of New Noto started off small and squat; to ward off earthquakes they were constructed above artificial “antiseismic” caverns. Within a few decades, however, all of the supposed antiseismic provisions had been forgotten and the New Noto had become a showcase for the finest Sicilian baroque churches and palaces. The replacement buildings were far from robust: several fell in the tremors of the next three centuries, and part of the cathedral collapsed after a rainstorm in 1996.33
Spanish colonial administrators had another chance to apply their strategy of relocation at Concepción in the Bay of Talcahuano on the stormy coast of southern Chile. On July 8, 1730, the town’s 200 buildings swayed in a distant earthquake, which was followed a few hours later by three massive sea waves that removed all the low-lying buildings by the port. Twenty-one years later, at dawn on May 25, 1751, a great wave emerged out of the shimmering darkness ten minutes after a heavy shock and destroyed the whole town. Chile’s viceroy announced that the city would relocate to a site seven miles inland, on the alluvial plains of the Mocha Valley on the Bio Bio River.34 Ten days later, a grid of streets had been drawn and work started on the crown offices and monasteries. Within a month, opposition to the move was being led by the city’s bishop, José de Toro y Zambrano.35 Even as the new city began to take shape, the ordinary fishermen and stevedores, refusing to move, were rebuilding their hovels at the old site, now named Penco. In September 1754, when the authorities attempted to evict the implacable residents, the bishop excommunicated the bailiffs. The bishop’s unwavering opposition ended only with his death in 1760. Today, in memory of that forced relocation, the citizens of Concepción still refer to themselves as “Penquistas.”
While the new inland site of Concepción was beyond the “tidal waves,” its alluvial soils offered no refuge from shaking. In 1835 the new city was leveled in a great earthquake, and it happened once again in 1939, when 80 percent of the 15,000 adobe structures in the city collapsed.36
The grandest eighteenth-century city to face relocation after a disaster was in Guatemala. Conquistadors founded the city of Santiago de los Caballeros de Goathemala on July 25, 1524. The original site was relocated twice in its first twenty years—first for defensive reasons and then again after it was wiped out by a volcanic mudflow.37 The new site prospered: it became the capital of Spanish Guatemala—a province that covered most of modern Central America—and Guatemala City was the third-largest in the Americas after Mexico City and Lima. On September 29, 1717, however, an earthquake ruined more than 3,000 buildings.38 The city’s leaders debated relocation, but instead the city was rebuilt in a bulky colonial adaptation of the latest European baroque style, with magnificent monasteries, nunneries, and churches embellished with stone carvings.
On the afternoon of July 29, 1773, the city of 40,000 was hit by another strong earthquake. Many people ran outside and were spared when the catastrophic (estimated Magnitude 7.4) main shock arrived fifteen minutes later.39 Every building was damaged, and many had collapsed. More than 500 people were crushed by the rubble.
Within a week, the governing captain-general proposed that the city be relocated. He created a commission to review the options while strongly recommending that the new site should be in the Valle de La Ermita, 15.5 miles (25 kilometers) to the east.40 It was initially claimed that the site was protected from earthquakes, but it was soon discovered that the two churches at La Ermita had been damaged in previous tremors. The commission chose to ignore this flaw and instead focused on the positives: the water was good, the earth was “solid,” and deep ravines through the La Ermita site would create escape routes in an earthquake. Later the ravines were declared “antiseismic.” After waiting more than a year for the slow exchange of letters, the choice of La Ermita was officially endorsed by the Spanish king.
In the new city, plazas were to be enlarged and streets widened, and just as at Lima after 1746, the maximum building height was set at 13 feet (4 meters). The new cathedral was to be fitted with wooden ceilings rather than vaulting. In 1776 two prototype “earthquake-proof” houses, like those at Lima, were constructed on the main square, in the hope that builders would copy them.
Terrified by the persistent aftershocks, initially people were enthusiastic about moving out of their ruined city. However, within a month the archbishop, Pedro Cortés y Larraz, and the city council, along with thousands of residents, were declaring that earthquakes occurred everywhere in their country and that the relocation proposal had been a panicked response.41 The administration was furious. In 1776 the captain-general sent his troops to force the inhabitants to leave, but five years after the earthquake there were still 11,000 left in the old ruined city. The Spanish king, Charles III, had the archbishop expelled from Guatemala for his “scandalous, offensive and contentious opposition.”42 Without their leader, many of the remaining inhabitants were evicted and their businesses forcibly closed, until the streets of the old city were almost empty. The poor returned surreptitiously to live among the haunted ruins of what was now known as Antigua (Old) Guatemala. By 1850, it had 9,000 inhabitants.
To abandon the city was to save it. Without its earthquake, this early eighteenth-century baroque masterpiece would have been overtaken by every progressive architectural and functional fashion that has so disfigured other cities in the region, from bottling plants to bus stations, from shopping malls to six-lane highways. Instead, its architecture was preserved, fissured and brooding, bookended by one earthquake that cleared the site for the city to be built and a second that froze it in time.43
At the relocated site of the new Guatemala City, with its “protective ravines,” little thought was given to shock-resistant buildings. Thus, earthquakes in December 1917 and January 1918 more or less destroyed the new city.44 The archbishop had been right.
Relocation also has a political imperative. A ruler can take money and power from an urban elite whose land values will drain away once economic activity has been diverted, while enriching the landowners of the new city site, who might, as at Noto, be political patrons. Following its calamitous 1944 earthquake, the national government proposed that the city of San Juan in western Argentina should be relocated, but the old city elites defeated the move, campaigning through their control of the local press.45
Another elite not prepared to be dislodged were the oligarchs of the port of Galveston, Texas. In 1838 a group of investors formed the Galveston City Company to begin the development of a seaside town and port on the barrier island. The author of Braman’s Information About Texas (1858) declared: “I should as soon think of founding a city on an iceberg as on Galveston Island, if I looked to its safety and perpetuity.”46 The town proved a huge success—by 1900, twenty-six millionaires lived in one five-block section of mansions. With its regular steamship destinations and consulates, the island had pretensions to being the Manhattan of the Gulf Coast.
However, the highest ground on the island was less than nine feet above sea level, while some buildings had thresholds only three feet above the tides. The city grew familiar with flooding in hurricanes.47 There were three floods in 1871 alone. On one occasion, a schooner and three sloops were abandoned on the streets. After the whole island was flooded in 1875, the city established a commission to consider building a seawall, but the proposal was rejected as “too costly.”
Then came the hurricane of September 8, 1900. Winds estimated at 145 miles (230 kilometers) per hour drove before them a 15-foot (4.5-meter) surge accompanied by colossal waves, turning the houses near the coast into a mass of floating timbers several stories high that acted as battering rams to scour buildings off their foundations six blocks inland.48 An estimated 8,000 people died.
That would seem to have been a good time to relocate the city. Yet some hefty real estate values were at stake.
In January 1902, the city appointed a Board of Engineers, who were tasked to raise and protect Galveston Island. By this means, landownership would not be compromised. The plan required 3,000 buildings to be jacked up by 11 feet (3.3 meters), sand pumped beneath them, and all to be protected by a 17-foot concrete seawall.49 The program of work took seven years to complete. The city never recovered its former status.
At the end of the twentieth century, urban relocation has come back into fashion after the 2005 earthquake in Kashmir, Pakistan, and the 2008 quake in Wenchuan, China, only now the scientific rationale is more convincing—to avoid catastrophic landslides.50 The policy remains no less controversial.
INSTEAD OF FACING THE UNPOPULARITY OF RELOCATING A CITY, why not redesign it on its original site?
Destruction provides a blank canvas on which an ambitious architect can redraw the cityscape. It was the opportunity to transform the narrow winding streets of medieval London that inspired Christopher Wren’s vision of a city of wide boulevards and grand piazzas after the 1666 Great Fire.51 Time and again after catastrophes, however, the obstacles to grand design have been the same: landowners need to be bought out to make way for new boulevards, but the devastation has removed a city’s income from rents and taxes.52 There were 80,000 homeless Londoners clamoring for action, and the city’s prestige as Europe’s third-largest city needed to be restored. Frustrated, Wren railed against the “obstinate Adverseness of a great Part of the Citizens.”53
Pragmatically the commission proposed a rebuilding act, which was passed by Parliament in February 1667.54 All property owners were forced to give up a few feet of their frontage so that roads could be widened. The original geometry of the city’s streets would reassert itself, like new saplings emerging on sawn stumps. By 1672, the city had been largely reconstructed on its old street plan.
Both the 1756 plans drawn up by the precocious twenty-seven-year-old Scottish architect Robert Adams for a neoclassical Lisbon and Daniel Burnham’s 1906 visionary redesign of San Francisco with Parisian avenues and boulevards foundered on the same obstacles that blocked Wren: a lack of money to compensate displaced landowners.55 It only proves possible to completely redesign a city’s plan after a disaster when the majority of the landowners are either dead or dispersed. When Catania lost 16,000 of its 20,000 citizens in the 1693 Sicily earthquake, the opportunity arose to ignore the “obstinate Adverseness” of the citizenry and completely reconfigure the medieval city’s plan with wide straight streets and new piazza squares (said to allow building owners to camp next to their damaged buildings in a future disaster).56
THE EARTHQUAKE OFFSHORE AT SANTA BARBARA, CALIFORNIA, AT 6:43 a.m. on Monday, June 29, 1925, had tumbled the facades of brick stores and hotels, ruining thirty-six blocks of the business district. Only the time of day limited the death toll to thirteen. The first significant earthquake in California since 1906 was generating a lot of bad publicity and threatening the vital Los Angeles tourist trade.
On the day after the earthquake, Santa Barbara community leaders met to form the Board of Public Safety and Reconstruction. As at San Francisco in 1906, the board was looking for some instant positive “spin” to counter the tales of death and destruction. There was no fire in which to shroud the earthquake. A wealthy engineer in the group, Bernhard Hoffman, successfully argued that the city should be rebuilt in the architecture he had practiced in the neighborhood of El Paseo, a village of mixed shops and restaurants imitating elements of an original Spanish colonial building known as the “de la Guerra 1849 adobe.”57 With a flair for publicity, he christened this architecture the “Santa Barbara style.” The announcement was headlined the following day in the local paper: “Spanish Architecture to Rise from Ruin.” The committee listed a number of buildings that contained elements of this style and had survived the earthquake unscathed. However, there was no real evidence that this style of building was any better than others at resisting shaking damage.
Within two weeks, the Board of Public Safety and Reconstruction had created the Architectural Advisory Committee and the Architectural Board of Review, on both of which Bernhard Hoffman served.58 “Now when everyone is absorbed in the story of our misfortune,” a spokesman for the board announced, “let us surprise the balance of California and the world by turning our misfortune into a source of rejoicing.” The earthquake had become the opportunity to impose a unified architectural aesthetic.
The idea of the “disaster makeover” quickly caught on. A devastating hurricane in August 1926 had stripped the tourist town of Miami Beach back to its foundations. In the reconstruction of South Beach that followed, architects designed whole blocks in Art Deco style.59 The best remedy for the bad publicity that accompanied a disaster was for a city to be resurrected in the latest fashion.
On February 3, 1931, the town of Napier on the east coast of North Island, New Zealand, was completely destroyed by a massive earthquake situated on a fault that directly underlay the city. Based on the precedent of Santa Barbara, and perhaps showing some “California envy,” a group of Napier architects campaigned to create a strong local Art Deco–influenced style and worked with the town’s reconstruction committee to achieve consistency in the appearance of all the replacement buildings.60 Out of fear of further earthquakes, the buildings were kept low—no more than two stories.
The success of the 1925–1931 aesthetic reconstruction narrative at Santa Barbara, Miami Beach, and Napier might be thought to have launched many more post-disaster imitators. Surprisingly, Napier proved to be the end of the road for this movement.61 After the next significant earthquake, in Long Beach, California, in 1933, the reconstruction agenda moved to concerns of greater gravity, making the idea of rebuilding a city in the latest architectural fashion seem a tad decadent.
IN THE EARLY EVENING SHOCK, MANY NEW SCHOOL BUILDINGS, with their large roof spans and open halls, had collapsed. Damaged school buildings dominated the front pages of the local newspapers, and the head of the Long Beach School Department estimated that 6,000 children could have been killed if the earthquake had happened an hour earlier.62 Don C. Field was the Republican state representative for the district covering Long Beach and was himself a building contractor. Within two weeks of the March 10, 1933, earthquake, he introduced a bill requiring that the design and construction of public school buildings be regulated by structural engineers acting for the state.63 The Field Act was passed by the California State Legislature on April 10, 1933—thirty days after the earthquake.
The 1933 earthquake brought down the curtain on the long-running, business-led campaign of American “earthquake denialism.” There were practical actions to be taken to withstand the earthquake threat—the mortal disease could be treated.
The Field Act focused on new construction, but if 10 percent of enrolled parents requested it, the state architect would evaluate any school building. Within a year, 333 applications had been submitted to review plans for new schools, while 1,000 had been requested by parents of children in preexisting schools. In 1939 the act was expanded to require inspection and, where needed, retrofitting for all pre-1933 school buildings. The act provided much-welcomed employment for California structural engineers throughout the Great Depression.
The proponents of California’s Field Act proclaim with pride, “No student, teacher, or anyone has been killed from an earthquake in a school building since its enactment in 1933,” but in fact, since 1900, no major Californian earthquake has occurred in the vicinity of a town or city during school hours.64 The idea of declaring zero tolerance for school casualties was a brilliant initiative, but curiously, no one sought to export it or to make it a global standard.
WITH MORE EXPERIENCE OF DISASTERS THAN ANY OTHER COUNTRY, Japan has a pair of characters in its language that translate as “building back better.” This is contrasted with a closely related pair meaning: “restoration to the state that existed before.”65 A catastrophe provides the opportunity to make genuine improvements to reduce the risk.
One of the earliest examples of “building back better” was London after its 1666 fire. Yet the imperative for rapid reconstruction often overtakes the ideal of “building back better,” as when San Francisco completed its rebuilding in seven years, unhindered by any building code mandating earthquake-resistance, or when, after Hurricane Ivan’s storm surge flooded almost the whole of Grand Cayman, buildings were reconstructed at the same elevation as before.
The first city to aim to take the time to achieve genuine “build back better” after an earthquake disaster was Messina, Sicily.66 Because Messina’s 1908 devastation was accompanied by high mortality and the subsequent dispersal of the residents, and because the city lacked the “get up and go” commercial imperative of a San Francisco or London, it was possible to contemplate a complete urban redesign. In the new plan, commissioned by the city’s planning director and approved at the end of 1911, the size of the city was doubled, and the streets were widened and placed on a grid.67 No buildings could be higher than two stories. The world’s first earthquake building code required that buildings be able to withstand a sudden sideways force one-tenth of the building’s weight.68 However, within little more than thirty years, Messina had been devastated by World War II bombs and shells rather than earthquakes; after the war, all precautions about earthquakes had been forgotten, and the buildings of the city were rapidly reconstructed four to five stories high.
WHICH COUNTRY HAS MASTERED EARTHQUAKES IN THE SAME WAY the Dutch have mastered floods? Which has come closest to curing its shaking catastrophes?
In February 1835, while resting on the ground in a wood at Valdivia, Charles Darwin had the formative pleasure of experiencing the largest central Chile earthquake of the nineteenth century. A few days later, he visited the second city of Concepción; shocked by the destruction, he reflected that no civilized country could sustain or tolerate this level of repeated devastation.69 We now know that Chile, situated alongside one of the world’s fastest-moving plate boundaries, is on the earthquake front line. At the start of the twentieth century, relative to its population, Chile had some of the highest earthquake casualties of any country. Most deaths were caused by the fall of traditional adobe buildings.
On February 27, 2010, a near-repeat of Darwin’s earthquake came around again: 250 miles (400 kilometers) of the plate boundary dipping down beneath the coast of central Chile broke in a Magnitude 8.8 earthquake—at the time the fifth-largest earthquake ever recorded. It was early Saturday morning, before sunrise, at the start of the last weekend of the summer holidays.
Oklahoman Will O’Donnell was sleeping on the fifteenth floor of an apartment building in Vina del Mar, on the coast north of the earthquake rupture.70 Awakened at 3:30 a.m., he got out of bed and as the shaking became stronger and more rhythmic found himself thrown from side to side, bouncing off the opposing walls of his room. The lights went out, his desk was toppled, and plates crashed onto the floor. After ninety seconds, the shaking wound down and he looked out over a darkened city, to the cacophony of car alarms and the mad barking of hundreds of terrified dogs. There was no cell phone service. He walked down the fifteen flights of the staircase and out into the street, where he found the owner with a flashlight, surveying the base of the building.
Across the city, damage was limited to some twisted and cracked concrete where tall buildings had thrashed on their foundations. Even though the fault started to break next to the second city of Concepción and pounded the forest of skyscrapers in the inland capital of Santiago, in all the furious shaking only one tall building fell—a Concepción apartment building, not yet fully occupied, was torn off its cavernous basements.71 Millions of people were in the zone of strongest shaking, but only a few hundred people died, half of them in the tsunami. Many of the lives lost on February 27 were the result of some oversight—like the campers on an estuary island with nowhere to evacuate to ahead of the nighttime tsunami, or those driving the Santiago toll road when the overpass ahead of them collapsed.
Fewer people died in building collapses in the Magnitude 8.8 central Chile earthquake than in the 1,000 times smaller 2009 L’Aquila earthquake in central Italy.72 After fifty years of investment in world-class schools of engineering, and by strictly enforcing tough building codes, Chile today has some of the safest earthquake-resistant building stock in the world.
Chileans have come to fear and respect great earthquakes in the same way the Dutch fear and respect storm surge floods. Just as there is no corner of Holland entirely free from floods, so there is nowhere in Chile you can escape from shaking. All Chilean politicians expect some disaster during their period in office. Earthquake culture runs deep. Building codes are respected and followed, and developers hold a ten-year liability for building damage, with the threat of a jail term.73 Builders know that their handiwork will be tested over the course of their working life and that this is a matter of life and death.
The response to the 2010 earthquake still left plenty of room for improvement. It was the job of the revered Chilean navy to provide a tsunami warning, but no one was on duty that night, and anyway, communication systems crashed with the loss of electrical power.74 There still needs to be zoning to prevent development in the tsunami flood zones and a funded replacement program for the dangerous old adobe housing. Nevertheless, one day there will be a great Chilean earthquake in which nobody dies.
THE CARIBBEAN ISLAND OF PUERTO RICO LIES AT THE EASTERN END of the Greater Antilles, a fertile and rugged landmass 100 miles (160 kilometers) long and 40 miles (65 kilometers) wide. Without any gold or valuable minerals, the island was neglected by the Spanish colonial administrators, but then wrested away by the United States in 1898. After some relatively calm decades, on September 13, 1928, a Category 5 hurricane (known after its saint’s day, San Felipe), with sustained wind speeds of more than 150 miles (240 kilometers) per hour, made landfall along the middle of the eastern end of the island and proceeded to move to the west-northwest over the capital San Juan.75 The winds obliterated the island’s fragile thatched bohios and one-room chattel houses, tore off the tin roofs of the estancias and parish schools, and brought trees crashing down on the hipped roofs of the town houses, leaving a tally of 25,000 houses destroyed and another 192,000 damaged.76 Half a million people—one-third of the population—were left homeless, and more than 300 were dead. Four years later, the island was hit again, this time by the smaller (Category 4) San Ciprian hurricane, which traversed the entire length of the island from east to west, taking another 250 lives and leaving more than 75,000 homeless.77
Following the second storm, engineers carefully recorded which buildings survived and which ones had failed. They found that “only the heaviest construction of masonry and concrete with cemented tile roofs, came out of the zone of heaviest damage unscathed.”78 Other concrete walls with insufficient cement, with too widely spaced reinforcement, or with roofs inadequately anchored proved lethal to the occupants as well as to those downwind. Corrugated iron roofs fixed with smooth or even twisted nails, many of them installed since the 1928 storm, “were carried off like so much cardboard,” while the same roofing material fixed with bolts to a properly anchored frame survived.
One catastrophe can be put down to chance. Two becomes a pattern. The construction lessons learned from San Ciprian, following in the wake of the devastation from San Felipe, came to transform the island’s building stock. By the 1950s, a combination of an on-island cement industry and the marketing of the American suburban, car-owning lifestyle had encouraged the middle classes to move into estates of reinforced-concrete “bunker” houses with tiny jalousie windows.79 Hot and confined, these houses were a total break with the traditional airy rural houses, but they were deemed “bulletproof” in the most intense hurricanes, as proved when in 1989 Category 4 Hurricane Hugo hit the island along almost the same path as San Felipe. Only nine people died.
On the island of Bermuda, the lessons had been learned from two fierce hurricanes at the beginning of the eighteenth century. Afterward, only single-story buildings with walls and roofs built out of stone were constructed.
In 1994 the Florida counties blasted by Hurricane Andrew—Broward and Miami-Dade—adopted a tough building code so that new buildings would be designed to withstand 150-mile-per-hour wind speeds. Following seven years of protracted negotiations with building industry lobbyists, however, the building code finally adopted for the rest of the state was much weaker. Even when the catastrophe was just a few counties away, the local risk was not deemed “intolerable.”
How to perform a scientific experiment? Get one half of an island to apply a building code and in the other half leave the builders unsupervised. Then run the island through an intense hurricane. The experiment was performed in 1999 when Hurricane Luis pummeled the French/Dutch Caribbean island of St. Martin/Sint Maarten. The French had a building code. The Dutch did not. Buildings in the French territory were generally unscathed, while the Dutch side resembled an Alabama trailer park after a passing tornado.80