Fire and Lies: Did the Berkeley Fire of 1923 Contribute to the Tunnel Fire of 1991?: Part 2: Lies

Within days of the 1923 fire, plans for rebuilding more than six hundred houses in Berkeley’s Northside district began apace. City building inspector Robert Grieg immediately announced he would no longer issue permits for wood roofs on new houses.² This was not a new idea. In the first century Nero banned all wood structures after a six-day fire leveled Rome.³ But in 1920s California, Grieg’s response was decidedly novel. Today many take for granted building codes that mandate fire-protecting materials and construction techniques, smoke alarms, and fire sprinklers, plus city regulations mandating defensible space and clean yards. In 1923 it was different. Of the 118 cities in the United States that had ordinances banning shingles, not one was west of the Mississippi River.⁴ After the San Francisco fire of 1906, fire-safe building codes were promulgated in the downtown districts, but none in the residential districts. This was typical. Building codes tended to focus on protecting mercantile districts and apartment houses from fire, and rarely if ever touched upon single-family residential dwellings.⁵ This was precisely the case with Berkeley’s Ordinance 779, the city’s Uniform Building Code, adopted in 1922. It banned all wooden structures within the limits of four fire zones.⁶ But none of the four zones was residential, and they comprised only a small fraction of city land.⁷ Beyond these zones there were no restrictions on exterior materials.

Seven days after Berkeley’s fire, the Oakland City Council passed a ban on shingles on all new construction.⁸ Prodded by UC professor of political science Samuel C. May, the Berkeley City Council began discussing a similar ban. When Berkeley Mayor Frank D. Stringham proposed specific language, the council took no immediate action and decided instead to hear from its citizens.⁹

With the fire out, Berkeley city manager John North Edy began asking what was known about the origin of the blaze. Upon receiving a slew of conflicting information he decided to convene a group of three experts to study the matter. He invited the state government, federal government, and University of California to appoint one member each to a new commission. The university assigned forestry professor Emanuel Fritz. The U.S. government assigned U.S. forester James W. Nelson, holder of a faculty appointment with UC Agricultural Extension. The State of California assigned Earl Barnes, who, as chief of the Contra Costa Hills Fire Protection Committee, participated in the early effort to suppress the blaze.¹⁰ This trio of experts was instructed to recommend plans for improving the watch, suppression, and “permanent prevention of the spread of hill fires to Berkeley property.”¹¹

Eleven days after the fire, Edy received his requested report.¹² It assigned the most probable cause to a careless smoker on a trail: “The fire did not originate at the road…no evidences of incendiarism [arson] were found, neither was there evidence that a broken or grounded power wire was responsible. The most probable cause [is the] carelessness of smokers traveling along the power line trail.”¹³ Their conclusion was at odds with what Jordan, the firefighter first on the scene, told multiple reporters—that he “saw the start of the blaze” and it was caused by “a high tension wire, snapped by the freak wind.”¹⁴ A later study by the National Board of Fire Underwriters made no statement as to cause,¹⁵ nor did California’s state forester.¹⁶ A 1924 lawsuit alleging malfeasance of the Pacific Gas and Electric Company (PG&E) was successfully rebutted. Thus the true cause of the blaze remains ambiguous.

To minimize the chances of a future wildfire, Edy’s experts called for a full-time lookout on a fire tower to be built on Grizzly Peak, and the clearing of a firebreak on the ridgeline. Their report said nothing about roofs with wood shingles. Also the report gave no information on what Barnes had done to fight the fire other than to note he was “unable to cope with the situation with the absolutely inadequate organization and means at his disposal.”¹⁷

The experts were not asked to comment on what, if any, bad policy decisions led to the feeble first response. If they had, they might have concluded, as did former California governor George Cooper Pardee, that the recent “economy plan” of Governor Friend William Richardson had so starved the financial resources of state fire fighters that ordinary citizens felt compelled to organize their own battalions, however ineffectual they might be.¹⁸ (Richardson also ignored Berkeley’s post-fire request for relief.)¹⁹

With report in hand, Edy immediately pushed for fire patrols, a watch tower, fire spotters, water towers, a plowed firebreak on the ridgeline, a northern extension of Skyline Boulevard (present-day Grizzly Peak Boulevard), and authority to negotiate on behalf of the council with the East Bay Water Company for improved water mains.²⁰ Without objection, he would soon get everything he asked for.

On the same day Edy received his report, the Berkeley City Council heard recommendations from seven fire experts and three shingle industry representatives. (The three industry representatives were Fred W. Foss, a lumber company owner and Alameda County treasurer; James Miller Hotchkiss, “head of Hobbs-Wall Company, one of the largest redwood lumber companies”; and Richard Fox Hammatt, secretary and head of the California Redwood Association. Four industry groups were behind the effort: the California Redwood Association; the “shingle branch of the Western Coast Lumbermen’s Association”; the California Sugar and White Pine Manufacturer’s Association; and the California Retail Lumber Dealers.)²¹ The fire experts all urged the council to approve a ban on shingles, whereas the industry representatives all loudly demanded the city prove in a court of law that shingles were less fire resistant than roof tiles surfaced with fine gravel. Despite conventional wisdom that wood burns and gravel does not, the industry representatives proclaimed that if the council wanted to enact an “unbiased” ordinance, it first had to establish a rank order of fire resistance of all roofing materials. Although city manager Edy sent a telegram to the U.S. Bureau of Standards requesting data on the industry’s central allegation that wood shingles were no more liable to burn than asphalt tiles surfaced with gravel, he would not receive a reply for two months. The bureau replied, saying the work was in progress and results would be in hand “within a few months.”²² The bureau had just published a pamphlet under the promising title Recommended Minimum Requirements for Small Dwelling Construction, but it too was of little help. Despite asserting that “the great danger of the wooden-shingle roof is from chimney sparks and flying brands from burning buildings or bonfires,” it contained no data, and “no direct requirements for roof coverings are included.”²³ Therefore, because roof standards were nascent or nonexistent, Berkeley city leaders were in a quandary; given the many eyewitness accounts given by Berkeley residents, city leaders must have seen with their own eyes the grave danger of airborne shingles carrying flames from house to house. Yet they were unable to muster hard data from a trusted agency to rebut the lumber industry’s nonsensical claims. For good measure, the industry threatened to sue.²⁴ Any attempt to limit shingle sales were, in their opinion, deeply objectionable.

Yet the lumbermen’s logic was faulty. We argue here that the relative order of flammability makes no difference. If wood shingles and some other roofing material, however fire resistant they may be, were both a menace, then both ought to have been banned. All that mattered was whether a given roofing material did or did not ignite when touched by flame. The council members knew wood shingles were a main cause of the size and speed of the recent conflagration, and they knew they had an inescapable obligation to ban them: Action was obvious and necessary.

On September 28, eleven days after the fire, the council unanimously passed an anti-shingle ordinance, one step short of formal adoption. It approved roofing materials of “concrete, brick, tile, terra cotta, slate, galvanized iron or tin” for “all buildings hereafter erected.”²⁵ Although the intention of the ordinance was obvious—to ban wood shingles—wood was nowhere mentioned except by inference in a clause that “permitted other fire resistive roofing materials…if such materials, roofings or shingles are substantially equally as durable and fire resistive as those that are approved and bear the label of approval of the Underwriter’s Laboratories.”²⁶ As residents of the burned district focused on the nitty-gritty task of rebuilding and refurnishing their homes, everything seemed to be on track at city hall to protect the town from a repeat of the conflagration. The Gazette praised the council’s action:

Ostensibly, Berkeley’s civic and private spheres had learned their lesson, were making amends, and were moving on. And so the story might have ended there, but it did not. If the lumbermen wanted to continue selling shingles to Berkeley homeowners, they needed either a standards lab to demonstrate wood shingles were “fire resistive” or “resistant”—which would never happen—or they needed to reverse the council’s decision. The industry’s demand for an unbiased ordinance would soon serve its purpose: to confuse the public about the palpable peril of urban wildfire and to allow the industry time to spin a new narrative based on grievance. If the matter could not be settled in a court of law, there was always the court of public opinion.

The new ban on shingles enraged the wood industry. It launched a fervent campaign to reverse the decision. There was still time. According to the regular order of council business, a bill is passed first and then after an interval it is adopted, at which point it becomes a legally binding ordinance. In the interval there exists an opportunity to persuade council members to change their vote. The wood industry lost no time in making the attempt. It sent a letter to the council arguing against the ban on the grounds it would impose a harmful financial burden on homeowners, which was inarguably true for the tiny fraction of Berkeley homeowners whose roofs needed to be replaced at that very moment, but whose financial affairs were so precarious that the additional cost of asphalt over wood tiles (if such an extra cost existed) was “harmful.” Also, the industry lobbyists falsely claimed that “shingle roofs constitute no more of a fire danger than do roofs constructed of [fire-resistant roof tiles].”²⁸

The lobbyists then turned to grievance. They ran the first in a series of large ads in Berkeley and Oakland newspapers. The headline asked, “Why pick on shingles?”²⁹ By this means the industry cast itself as a weakling and underdog. Furthermore, the ad accused the council of “subterfuge,” a classic tactic of accusing one’s opponent of one’s own misdeeds.

A day later the pro-shingle forces published their second big ad. Its title characterized efforts to ban shingles as “born in a moment of hysteria.”³⁰ In 1923, the word “hysteria” implied a predisposition by females to excessive or unwarranted nervousness. Even if not intended to convey a certain coded misogyny, the ad’s hysteria claim was an overstatement. Of course, there had been widespread fear and maybe even a few instances of hysteria during the fire, but all that vanished once the fire was out. Yet the ad pushed the false notion that opposition to shingles was now fundamentally irrational. This was just the beginning of what James Thurber once described as “the havoc wrought by verbal artillery on the fortress of reason.”³¹

Soon after, the Gazette reported that industry lobbyists had “bombarded the mail of every fire sufferer, regardless if he were a house owner or not, with letters, editorials and posters calling for a ‘square deal’ for wooden shingles.”³² Thus the lumber industry promoted the idea that it was an aggrieved victim up against some sort of faceless bully rather than a group of duly elected officials obliged to protect the public against fire.

The tactics of the lobbyists were well rehearsed. One year earlier a referendum had been brought before the state’s voters. Although this proposed “Housing Act” contained a mass of overwhelmingly popular Progressive Era reforms, it contained a ban on wood shingle roofs, and thus the entire measure was targeted for defeat by the lumber interests. Under their influence, an attack was waged by the industry’s lobbyists. Soon, even the newspapers were routinely referring to the measure as the “Shingles Act.” It was rejected by voters by a large margin. A year later, when stripped of the shingles ban, the measure easily passed in the legislature and has since evolved into the state’s standard building code.³³

On October 9, 1923, the Berkeley City Council met to cast their final vote on the shingle ordinance.³⁴ Before a vote was taken, a representative of the wood industry presented a petition with four hundred signatures protesting the anticipated financial burden to homeowners if the ban was approved. In putative opposition to the protest was councilmember Elmer E. Nichols, an Oakland attorney, who was interested “in making roof requirements more strict,” to which end he requested the council delay its vote. His request was granted.³⁵ But just five days later Nichols had not only changed his position, but he presented to the council an alternative draft ordinance authored by attorneys working for the wood industry. This alternative, if adopted, would have the effect of permitting unrestrained use of wood shingles in all new construction or refurbishing projects outside of the city’s mercantile district. The alternative draft was endorsed by a second member of the council, John W. Caldecott, a Berkeley pharmacist, and, inexplicably, by eminent UC professor Emanuel Fritz, who had previously served on the city manager’s commission for investigating the fire. Although Fritz considered the council’s original ordinance to be “by far the better one,” he suggested the council “make concessions” and adopt instead the wood industry’s proposed ordinance because it gave “a maximum of protection at the least cost.”

In the face of these unexpected reversals, councilmember May asked for a prompt vote on the council’s original ordinance. His request was granted, but only after another two-day delay.³⁶ When the city council finally did cast its vote, it was unanimous in confirming the original ban on shingles.³⁷ Although the vote count was a perfect defeat for the lumbermen, they were emboldened by the result, because two members of the council, Nichols and Caldecott, said they voted as they did only to eliminate the possibility of the city being forced to pay for a special off-cycle election that the lumber industry had threatened to demand, and that would have been required by law had the two not voted as they did.³⁸ Afterward the lumbermen privately congratulated themselves for having “split” the council, and they vowed to redouble their efforts to reverse the ban.³⁹ They did not dally. Within two days rumors reached city hall that the lumber industry was “determined to force an initiative”—meaning, introduce a ballot measure that, if passed, would rescind the council’s decision.⁴⁰ These rumors would soon prove true.

Working against the lobbyists’ efforts were fresh memories of houses bursting into flame, of homeowners on roofs with garden hoses and wet burlap bags, of houses on the edge of the main burn area surviving only because their sides were stucco or concrete. The homeowners’ common sense blamed wood roofs, and the experts concurred. A report released mid-November by the National Board of Fire Underwriters concluded the main cause of structure ignition was flying embers landing on wood roofs or wood sidings: “This fire is another to be added to the long list of those in which wooden shingle roofs permitted the spread of fire from building to building, thus, in conjunction with a high wind and a weak water system, resulted in a conflagration which destroyed a closely-built residential section and endangered much of the other part of this valuable territory.”⁴¹ Underwriters also gave a statistical argument to buttress its claim that shingles were not just a major cause, but the leading cause:

While the factors working against the shingle lobby were substantial, three factors worked in its favor. First was “some cooling of the popular demand for fire protection since the few weeks immediately following the fire,” according to the Gazette.⁴³ Second was what researchers at the University of Iowa discovered in 2013 to be “the initial optimism of the people living in communities with daily reminders” of a disaster they had personally witnessed. The researchers compared the attitudes of people who survived a devastatingly fatal tornado to those living in a nearby unaffected community. They found residents of the affected community significantly more optimistic than residents of the unaffected community. Why? Survivors were relieved they had “dodged a bullet” and comforted by the notion that “lightning doesn’t strike twice in the same place.”⁴⁴

The third factor favoring shingle promoters was that, unlike flames, wood is not perceived to be threatening. As author George Marshall has pointed out, “once things are accepted into our status quo and assumed to be part of normal life, it requires a far higher level of threat to have them removed…people are initially traumatized but dust themselves off and focus on reconstruction and moving forward.”⁴⁵ At about this time there circulated in Berkeley an anonymous and undated pamphlet, The Story of the Berkeley Fire.⁴⁶ Wrapped in an attractive red cover, it contained a panoply of fire photos that purportedly constituted “proof irrefutable of the falseness” of claims that “the fire was attributable directly if not solely to the use of shingles.” Almost certainly the pamphlet was published and distributed by the shingle industry.

The pamphlet’s first photo showed the incinerated remains of “the first house to be destroyed,” Professor Brodeur’s (Figure 1). (It may be true that Brodeur’s house was first to burn down, but there are at least ten rivals for that wretched distinction.)⁴⁷ Just beyond the debris in the foreground is the empty draw of Codornices Creek. On the left are a few indistinct roofs, perhaps three, and a house under construction or reconstruction. The shady side of the draw is covered in trees. A lone chimney on the ridge beyond is marked by a printer’s arrow pointing to the only visible remnant of Ethel Underhill’s house, her chimney. The photo was printed opposite two affidavits, one by Brodeur declaring his house was the first to burn down, and one by Underhill declaring hers was the second.

The photo and affidavits were published for the purpose of posing a question and giving a shingle-exonerating answer. “If shingles blew from [Brodeur’s] ‘key’ house,” the pamphlet asked, then “why did all shingle roofs in [the] draw escape undamaged? They were directly in path of flames, but clearing saved them from brush fire.”⁴⁸ Thus was it insinuated that if a shingle house in a clearing could survive the wildfire, then all shingled houses must be blameless for the conflagration: fault must lie instead with those who failed to clear vegetation from the perimeter of their houses. This was an early expression of the idea, well known today, of “defensible space”: to prevent a house from catching fire, the most important task is to clear vegetation away. Although promoted heavily by firefighters and elected officials in recent decades, the reality is that defensible space is necessary but insufficient alone to stop a structure from taking fire.

Consider a neighborhood built entirely of wood-roofed houses where the policy of defensible space has been taken to the extreme, meaning all vegetation has been removed, and the neighborhood is surrounded by a firebreak of great depth. No matter how a house in the neighborhood catches fire—whether by arson, human blunder, or natural causes—when it does, in a stiff wind, the impact is likely the same: The index house spews embers, houses downwind catch fire, and the process cascades. The flames are airborne, not ground-borne. They can, as in 1923, cause spot fires two miles distant from the main burn area. During the Oakland–Berkeley fire of 1991, one pilot reported flying past a burning shingle at 2,000 feet.⁴⁹

The industry’s pamphlet distributed in 1923 may have been exemplary in its visual appeal and in highlighting the importance of defensible space, but it failed to give photographic proof that shingles were not to blame for the speed and size of the conflagration. Were the houses in the photo in fact downwind of the index house? Were their roofs actually shingled?⁵⁰ Did anyone take action to protect those houses, as was known to be the case for at least three others nearby?⁵¹ Moreover, randomness is characteristic of wide-ranging fires, and Berkeley’s was no exception. Always a few wood structures prone to combust in fact survive and a few structures likely to survive are destroyed.⁵² The shingle industry did not address any of these questions, nor did the remaining ten pages of the pamphlet give any additional “proof” to exonerate shingles. It was nothing more than a show of the propagandist’s standard technique of card-stacking, the declaration of statements favorable to one’s argument—whether factually correct or not—and silence on all unfavorable arguments.⁵³

By December 5 the shingle lobby had organized and launched a second petition drive.⁵⁴ The goal was to collect 1,994 signatures of registered voters, amounting to 15 percent of the votes cast for mayor in the previous election. If obtained, those signatures would force a citizens’ vote on whether or not to rescind the council’s shingles ban. The wood industry hired eight signature collectors, all “well paid.” It was reported that, of those approached on the first day of collecting, 94 percent agreed to sign.⁵⁵ The focus was on getting signatures from residents of South Berkeley and West Berkeley, far from the main burn area.⁵⁶ Meanwhile, the Gazette scoffed that “the great majority will affix their signatures to almost any document.”⁵⁷ “Commonsense,” the editor opined, “should teach us not to compromise with the fire, but commonsense and human selfishness will not always work together.”⁵⁸

Two days after the drive began, Mayor Stringham advised voters not to sign the petition. He argued for fact-based policymaking, and for waiting for the results of tests promised by the Bureau of Standards. He noted that “in the face of the experience of this and other cities, it seems to be very unwise for citizens to sign initiative petitions attempting to legislate on this subject. Particularly is this so since the people most actively engaged in advocating the permissive use of wooden shingles are the lumber dealers themselves.”⁵⁹

Lumber company owner James Miller Hotchkiss wrote a rebuttal of Stringham’s comments. Primarily he criticized the National Fire Underwriters Association for “its claims that shingles are a fire menace, although it has never tested them.”⁶⁰ Despite the falsity of the claim that shingles were not a menace, it didn’t take long—just eight days—for the wood industry to announce they had collected the necessary signatures to force a vote of the citizens.⁶¹

Through the end of the year baseless assertions of the shingle lobby continued to influence the thinking of many, such as Berkeley resident Joseph A. Gilman. As former president of the Board of Fire Underwriters he ought to have been a reliable authority. At a public meeting in mid-December, however, he reminisced that “in the good old days shingles were shingles. They were split from good sound timber, they were longer and thicker at the butt and would withstand heat to a marked degree.”⁶² This was nonsense. It may be true that, in the past, shingles were longer or thicker or split from better timber, but none of that affected combustibility in any significant way.

Even Mayor Stringham began to fall under the spell. In December he informed the press he was interested in the upcoming results of tests by the U.S. Bureau of Standards “as to the fire-resistant qualities of wooden shingles.”⁶³ Rather than take the opportunity to denounce the malarkey being pushed by the lumber lobby, he entertained the possibility that the bureau might discover wood shingles were fire resistant. Nor did he mock those who pushed the idea that wood might prove to be fire resistant if only it was properly tested.

Perhaps out of fear of appearing to be biased, the Gazette reported Stringham’s remarks without stating the obvious, that it was preposterous for the city’s top elected official to be musing about the fire-resistant properties of wood.

By the start of 1924, everything was proceeding exactly as the industry wanted. First, signatures had been delivered to city hall—4,717 to be exact—enough to force a “voter initiated” ordinance be added to the ballot at an upcoming election.⁶⁴ (In fact it was not voter-initiated but industry-initiated, and probably would never have made the ballot if the lumbermen had not hired and paid signature gatherers.) Second, the vote would determine if the industry’s preferred ordinance would replace the city’s. Third, the industry’s alternative ordinance was written to its own specifications, by its own lawyers. Although their preferred language was nearly identical to the city’s, the net effect of a few small changes was to reverse the council’s ban on shingles. Both documents were technical and lengthy, thus likely to be read by no more than a handful. As dull as the exercise may have been, it would have made sense for someone—perhaps a Gazette reporter—to make a word-for-word comparison of the two documents. Proposed changes to the existing ordinance would have provided insight into the industry’s desire to suppress information about the danger of shingles.⁶⁵ One section, for instance, specified the conditions in which roofing materials were approved for use. In the council’s original ordinance, a given roofing material type was automatically accepted if it had the seal of approval of the national standards organization Underwriters Laboratories (UL). The wood industry switched this condition for their own, that the material would be automatically approved if it was “as hereinbefore prescribed,” meaning wood. The industry wanted nothing to do with UL or, for that matter, any other unbiased standards group, surely none of which would have concluded the absurd: that wood shingles were fire resistive, let alone fireproof.

The matter of a costly special election continued to linger. In late January industry lobbyists appeared before the city council to request repeal of the 1923 ban on shingles, arguing that since the council was bound to lose, it ought to concede in advance to avoid wasting the taxpayers’ money on a special election. Councilor May responded by characterizing the industry’s intentions as a “disgrace,” asserting it was “not interested in fire protection but simply to sell shingles.” In rebuttal the industry was defended by their chief spokesman Richard Fox Hammatt, head of the California Redwood Association, their Harvard-trained “publicity director.” As secretary-manager of the California Redwood Association from 1921 to 1931, Hammatt was responsible for national and local advertising. It was a job for which he appears to have been well qualified, having previously been employed as chief of the “information bureau” of the United States Forest Service, engaged in “public relation techniques to conserve forests,” meaning fire prevention.⁶⁶ Hammatt repeated the usual argument that it was discriminatory for the council to ban shingles without banning other roofing materials deemed inferior. He produced a document by the Wood Products Laboratory of Canada that purported to show asphalt tiles were less fireproof than wood shingles. May scoffed at the report, noting the lab was “maintained by the lumber interests of Canada.”⁶⁷ Unconstrained by truth, Hammatt would continue to orchestrate a campaign aimed at persuading Berkeleyans that shingles were not a menace.

With only six weeks to go before the vote, lobbyists for the wood shingle industry staged a demonstration that likely prompted many Berkeley citizens to vote against the council’s ban. The occasion was yet another of many public debates on the topic. Industry representative Hammatt brought forth what he claimed was an ordinary, off-the-shelf fire-resistant shingle, put a match to it, and set it ablaze. He announced that these asphalt tiles “burn quicker and longer than wooden shingles.” He contended, without evidence, that “the flying brand hazard was not eliminated by their use because they would fly almost as fast and with greater danger from fire than wooden shingles.”⁶⁸ What more was there to say?

In fact, each asphalt tile has two distinct sides. One side is laid face down, toward the roof deck, and the other side face up, to the weather. To impart fire resistance the weather side is covered with fine bits of gravel. Almost certainly Hammatt made no attempt to ignite the weather side, only the face-down side. But even the face-down side was fire resistant. It contained a layer of some fibrous material, such as felt or asbestos, intermixed with asphalt to create a water-impervious and fire-resistant layer. When these tiles are properly installed, the roof is protected against flame and heat. Thus, Hammatt’s demonstration was astonishing because it seemed to confirm, against all known facts and expert testimony, what precisely the wood industry had been claiming, that fire-resistant tiles were combustible. It was an apparent miracle.

If Hammatt’s demonstration was a trick, then someone—at least the Gazette—ought to have tried to expose it. Had the burning tile been presoaked in kerosene? Or, as a Berkeley resident would assert a few weeks later, had it been split open, laced with potassium nitrate, and resealed?⁶⁹ The Gazette never followed up, not even to query experts—of whom there were plenty on the Berkeley campus—on whether Hammatt’s demonstration was rigged and, if so, how. The reticence of the Gazette and its reporters is puzzling. A stream of misinformation was being pushed by the wood industry and its sham front group, the People’s Shingles Club. Misinformation was the main news. Even if the Gazette’s newsroom somehow forgot their journalistic obligations, one might expect the publisher to recognize the virtue of selling newspapers to readers eager to read an exposé of professional trickery.

In response to Hammatt’s demonstration, Councilor May might have promised to debunk the demonstration, or at least argued that if both roofing material types were combustible, then both should be banned. He did not, opting instead to discuss the statistics of the shingle hazard that, although compelling to the academic-minded, was unlikely to make any difference to voters who had decided the government was the problem, not shingles, as the industry was telling them.

A three-person debate on the shingles ban was held at the beginning of April. Frank V. Cornish—a Berkeley resident and former attorney for the City of Berkeley—appeared as a representative of the People’s Shingles Club, and said his organization was backing “the ordinance of the lumber interests.”⁷⁰ When pressed by May, Cornish conceded that many of the small fires in Berkeley every day were caused by sparks falling on shingle roofs. Cornish casually admitted that, from his point of view, “we have been getting along under these conditions for years, have been putting the fires out, and I guess we can continue to do so.”⁷¹ Memories were fading fast of “burning shingles sailing over the rooftops throughout an area a half-mile [square].”⁷² To the weary, there was a certain appeal to Cornish’s call to just ignore the whole thing.

On May 6, 1924, in thin turnout—eight months after fire leveled a Berkeley neighborhood—voters repealed the council’s ban on shingles. Fifty-eight percent of voters favored repeal. More than two-thirds of the precincts favored repeal, mostly in South and West Berkeley, those furthest from the fire (Figure 2).

For the wood industry, the result was gratifying, perhaps elating. The lumbermen could now count on sales of shingles to continue for the foreseeable future, not just in Berkeley but throughout the state. For the burned-out homeowners of Northside the election did not alter their plans. Despite having regained the legal right to construct wood roofs, virtually all chose to reconstruct with fireproof or fire-resistant materials. In an echo of the beauty of Santa Barbara, red roof clay tiles and white stucco siding began to appear.⁷³ Local architects were not merely accommodating but enthusiastic about the preferences of their clients. Bernard Maybeck initially championed the construction of houses with walls made of gunnysacks dipped in a cement slurry. He built one for his family of four.⁷⁴ Bubblestone never caught on, however, so his paying clients got traditional fire-resistant materials of construction. Although the master architect continued to enjoy a brisk practice, it was the end of an era. For the remainder of his career he designed only one more wood house.⁷⁵

Elsewhere in the city, new homes continued to be built with shingle roofs, and existing homeowners continued to replace old shingle roofs with new shingle roofs.⁷⁶ But the downside was looming disaster, particularly for homeowners whose houses survived intact. Frank and Ruby Gaines were one such couple. Having bravely, if foolishly, protected their all-redwood house when the flames swept past in September 1923, they chose to make no post-fire alterations. For more than fifty years the Gaines house, designed and built in the primitive vernacular of Charles Keelers’s Simple Home, would remain in its pristine as-built state.⁷⁷

Supporters of the shingles ban gave virtually no public response to their defeat. The only known printed comment was one of scorn by the editor of the Gazette:

Neither during nor after the campaign were there any allegations that the wood industry and its lobbyists had done anything illegal; only that, under cover of the First Amendment to the U.S. Constitution, they had violated the normative standards of ethical behavior by spreading blatant falsehoods. What little there was to complain about was that the lobbyists had been favored by toothless press coverage and incompetent counter-messaging. The Gazette was particularly at fault, having done nothing to expose the industry’s lies and deceptive tactics. Moreover, having never admitted this to its own readers, it never presented an explanation of how thousands of Berkeleyans were persuaded to overrule their elected leaders and vote against the ban on shingles. The lobbyists did not subvert the will of the voters through ballot stuffing, voter suppression, gerrymandering, or judicial decree. It was a straightforward campaign battle. Everything happened according to democratic rules that everyone knew and agreed to in advance. The bad guys won fair and square.

Despite having suffered defeat, the losers accepted the will of the electorate. Undoubtedly they hoped to reverse the decision, but for the immediate future there was nothing to do except wait for the inevitable return of the next urban wildfire.⁷⁹ Berkeley’s new predicament was described years earlier by a captain of the U.S. Army, James H. Simpson: “Facts are stubborn things, and he—be he engineer, statesman or philosopher—who ignores them will at length find that he has been following but a vain conceit which will eventually land him…into a condition of vain inanity or, it may be worse, of utter ruin.”⁸⁰

In September 1970, Berkeley experienced the first of two close calls. As a stiff, dry wind from the northeast blew through the Bay Area, an arsonist started a fire on Fish Ranch Road near the pass between Berkeley and Orinda. Thirty-seven houses were destroyed in the Hiller Highlands district of the City of Oakland. No one died.⁸¹

The second close call occurred ten years later, December 1980, when a fierce wind blew down a power line at Wildcat Canyon and Woodhaven roads and sparked a wildfire. Five homes were destroyed. Again, no lives were lost.⁸² Gary Cates, then Berkeley’s assistant fire chief for suppression, recalled feeling “lucky that we only burned up five houses.”⁸³

In 1981 the East Bay Regional Park District convened a committee of experts to “study the potential fire hazard existing along the twenty-five-mile boundary area between the north end of Wildcat Canyon Regional Park in Richmond and the south end of Anthony Chabot Regional Park in Oakland.” The committee was tasked to write “a comprehensive fire prevention program.” After seven months of work, the committee’s report and recommendations were adopted by the board of directors of the East Bay Regional Park District. The main recommendation was that a joint powers agency representing all municipal and corporate stakeholders be established to fund and oversee the creation of a twenty-five-mile-long fuel break in the Berkeley Hills, from Richmond to the southern limits of the City of Oakland. None of the committee’s recommendations touched on the importance of fire-resistant structures. The report’s only comment concerning fire-resistant structures was on a page headlined “Fire Prevention Retro-fitting Ideas for Homeowners.” The last bullet item in a list of seven “cautions to be taken for homes bordering on wildlands…for existing homes” was to “re-roof with fire resistant roofing materials.” This “idea for homeowners” was not repeated in a subsequent bulleted list of cautions “for new homes.”⁸⁴

For years the menace of shingles bedeviled Southern California, where the shingle industry reportedly made 75 percent of its sales. The region’s Santa Ana winds regularly triggered structure fires in residential neighborhoods. When a fire in the Puente Hills neighborhood of Los Angeles destroyed or damaged thirty-five structures in 1989, the city council immediately passed an ordinance banning shingles. The Cedar Shake and Shingle Bureau, a trade organization, protested the vote and threatened to sue the council, claiming their vote was “unwarranted and discriminatory.”⁸⁵ A year later, a national magazine dedicated to the renovation and maintenance of older homes published a theme issue announcing “the last word on wood shingle roofs.” None of the three articles on the topic mentioned any hazard.⁸⁶

A month later four fires in Santa Barbara and Los Angeles counties destroyed 637 houses and injured forty.⁸⁷ Houses with shingles constituted a disproportionate percentage of the structures lost. The wood shingle industry asserted that “treated wood roofs are not a fire hazard” and threatened to sue both the city and county of Santa Barbara for considering bans on shingles.⁸⁸

Five months later the wood industry ran a four-page paid advertisement complaining that “cedar shakes and shingles have been the scapegoat in residential fires throughout the state. Each time a structure with a shake or shingle roof is involved in fire, members of the press and some in government are quick to label these roofs as dangerous for California homeowners.”⁸⁹ The main argument of the ad was that “today’s cedar shake and shingle roofing products are not the same products of the past, and should not be subjected to the same condemnation.” It called out by name Los Angeles Fire Chief Donald Manning for continuing “to claim there is no evidence that pressure impregnated shakes and shingles retain their fire-retardancy for life in Southern California’s unique climate.”

In early 1990, Cates was promoted to Berkeley fire chief. He decided to reduce the risk of wildfires within city limits.⁹⁰ With the help of councilmember Shirley Dean, a report was brought before the city council February 1991 on the prospect of “prohibiting wood shake roofs in hazardous fire zones.”⁹¹ Four months later the council passed an ordinance “to designate a hazardous fire area and require fire retardant roofing therein.”⁹² The shingle industry made no public effort to counter the ordinance. Thus began the decades-long task of enforcing restrictions on roof replacement and hardening entire neighborhoods against urban wildfire. Delayed by lobbyists and their benefactors since 1924, the task was better late than never. The fire department also began enforcing a new requirement that homeowners in a newly designated fire hazard district in the hills maintain defensible space around their homes.⁹³

On September 17, 1991, Berkeley counted sixty-eight years without a major urban wildfire. A month later the city’s luck ran out, as did the luck of the City of Oakland.

On Saturday, October 19, 1991, a grassfire ignited in the vicinity of a house in Oakland’s Hiller Highlands neighborhood, just north of where Highway 24 enters the Caldecott Tunnel. The new neighborhood was noted for its spectacular views of the bay, proximity to San Francisco, and attractive homes. But the neighborhood was at obvious risk of a wildfire. An earlier report noted that many of the homes that had been built along the crest of the west-facing slopes had shake and shingle roofs and were surrounded by brushy vegetation that had been allowed, in many cases, to grow up to the homes and under decks. When people clear brush around homes, they rarely heed the thirty-foot minimum clearance required by state law, and even more rarely, the hundred-foot clearance that should be maintained in fire-risk hillside areas. Additionally, many of the streets in the area are narrow and winding, hampering access for fire apparatus and escape routes for residents.⁹⁴

The fire spread on a steep hillside just a few hundred feet below the ridgeline of the Berkeley Hills.⁹⁵ Both Oakland and the East Bay Regional Park fire departments responded. Five acres of chapparal were burned before the last of the visible flames were put out. There followed an extensive mop-up operation to unearth and douse embers smoldering in the thick layer of leaves and plant litter that covered the soil.

Mop-up continued early the next morning, Sunday, October 20. But as Oakland and East Bay Regional Park firefighters worked to extinguish hot spots, the wind picked up. At about 10:50 a.m., a gust caused the hillside to re-erupt. Thus began the Tunnel Fire.⁹⁶ An urgent call went out for a helicopter to drop water and fire companies to join the fight.⁹⁷ A home on Buckingham Boulevard burned down, the first of nearly four thousand.⁹⁸

Within thirty minutes of the first alarm, a falling power line struck and killed 49-year-old Battalion Chief James Merle Riley, as well as the 37-year-old woman he was shielding, Kimberly Dakis-Robson.⁹⁹ “Oakland’s command structure collapsed in shock for a while,” according to journalist Margaret Sullivan.¹⁰⁰ At about the same time, Oakland police officer John Grubensky, 32, having already directed to safety a reported one hundred residents, was escorting five more individuals in four cars down a narrow stretch of Charing Cross Road. The convoy was blocked and was then overrun by flames. Everyone died.¹⁰¹

By some measures the fire was similar to that of 1923. There was the same thick smoke and dark sky, the same pell-mell flight of residents as spot fires ignited ahead of the main fire line, the same tornado-like fire whirls erupting occasionally. However, in three aspects, 1991 exceeded 1923. One was the rate at which houses were destroyed: In the first sixty minutes, thirteen houses burned down every minute, on average.¹⁰² Second, unlike 1923, the 1991 fire was so intense, on terrain so steep, that it generated a self-ventilating firestorm. “My husband was doing a room addition at the time,” recalled Valarie O. Anderson. “The rebar in the foundation melted, as did a fireproof safe. One person I counseled saw a man spontaneously combust as he fled.”¹⁰³ And third, unlike the porous grid of streets in Northside in 1923, there were in 1991 (and remain today) only eight vehicular escape routes from the two neighborhoods struck first, Hiller Highlands and Grandview Canyon.¹⁰⁴

Communications channels overloaded, including one that caused Berkeley’s Chief Cates to be late arriving to his command center.¹⁰⁵ The deluge of calls to Oakland’s command center impeded the dispatch of orders to firefighters on the ground and in the air.¹⁰⁶ Visibility was nil. Crews ordered to work in unison had difficulty staying in contact. “One imagines Civil War soldiers, strayed from their company—rather than on a modern fireground,” writes Margaret Sullivan.¹⁰⁷

The fire burned steadily in the direction of Hotel Claremont and, by 3 p.m., it drew near.¹⁰⁸ Fire crews from Berkeley, Hayward, Oakland, Piedmont, and San Francisco departments focused on protecting the mammoth wooden structure, fearing that, if ignited, the inevitable inferno would launch an unprecedented spray of embers and create havoc for the as-yet-untouched neighborhood just downwind, Rockridge.¹⁰⁹ The fire was poised to grow into something far more dangerous.

And then at about 5 p.m. the wind changed, as it did in 1923. Hot gusts from the northeast gave way to a cool, moist, gentle breeze from the bay. Within a few hours most flames were out. Although spot fires continued to flare inside the main burn area for days, the perimeter was contained.¹¹⁰

It took a miserable week to find and identify all the dead. There were twenty-five. Eight individuals perished in their homes, three in their driveways, three in their backyards, and the remaining eleven on one of two streets, Charing Cross and Bristol.¹¹¹ The oldest victim was 85, the youngest a pair of 18-year-olds.¹¹² The coroner also reported thirty-four non-fire human decedents, presumably victims of heart attack and other afflictions triggered by the episode.¹¹³ The number of lost houses and apartments was a staggering 3,810—3,354 single-family dwellings and 456 apartment houses.¹¹⁴

In response to the 1991 fire, Oakland’s city council passed an ordinance requiring fire-retardant roofing material on all new construction, and on all existing structures in a newly designated fire hazard area.¹¹⁵ The State of California enacted similar legislation one year later.¹¹⁶ In a meticulous quantitative study published in 1994 by Napa County’s then–fire marshal, Ethan Foote, and UC Berkeley professor of forestry Keith Gilless, fire-resistant construction was ranked at the top of the must-do list to prevent or minimize urban wildfire.¹¹⁷ It is worth noting that, similar to the person vaccinated against a virus, a fire-resistant house is slow to catch fire, and unlikely to communicate fire to others. Thus, it neither risks neighbors nor consumes the time and energy of first responders. If a neighborhood is mostly built with fire-resistant materials, there exists the wildfire equivalent of herd immunity.

Why did Oakland and Berkeley experience a hellacious fire in 1991? We can say with certainty it was the result of a successful campaign waged sixty-eight years earlier by the wood industry and its lobbyists to block the Berkeley City Council from banning wood shingles. From that point forward, it was only a matter of time until, as Cates wryly noted after the 1991 fire, “the various elements circling in the cosmos line up—temperature, humidity, and wind. When they get lined up and a fire occurs, it’s ‘Katie, bar the door’!”¹¹⁸

Facts are stubborn things. Ignoring them can lead to ruin.

We thank the many librarians and archivists who have given freely of their guidance, particularly Berkeley Public Library (James E. Findley); City of Berkeley, City Clerk Department (Sarah K. Bunting); Oakland Public Library, History Center (Emily Foster); California State Library, Witkin Law Library (Garrett Shields); and University of California Berkeley, Bancroft Library (Susan Powell, Susan McElrath).

We acknowledge the importance to this research of online digital resources hosted by the Bancroft Library; Berkeley Public Library; news.google.com/newspapers; hathitrust.org; Internet Archive (archive.org); and California Digital Newspaper Collection, Center for Bibliographic Studies and Research, University of California Riverside (Brian Geiger).

For editing we thank Valarie Olsen Anderson and Karen Hansen Sibbett. For personal recollections and guidance we thank Valarie Olsen Anderson and Gary Leon Cates.