Two weeks before the first trial over claims that a tainted fungicide wiped out thousands of acres of crops and plants, three defense attorneys for DuPont Co. huddled for a conference call.
On the line were two researchers from a lab hired to test soil sprayed with DuPont's Benlate DF fungicide. The initial results, the researchers reported, showed traces of precisely what the growers contended had destroyed their nurseries: one of DuPont's deadliest weedkillers.
But DuPont never turned over to the plaintiffs those early results and subsequent tests supporting the growers' claims, as it had agreed to do and as court rules appeared to require. Instead, DuPont's attorneys presented the court with a summary of the tests, one they said proved there was no contamination.
Until recently, this courtroom aggressiveness seemed to pay off. Fearing defeat, the four nursery owners who brought the case agreed to a bargain-basement settlement. And across the country, 300 other growers soon followed suit.
But now DuPont's strategy could be backfiring. Those early test results and other raw data surfaced in a separate Benlate trial in Hawaii last year.
The federal judge who tried the first case _ known as Bush Ranch after the lead plaintiff _ ordered DuPont to explain why it shouldn't be found to have committed fraud by withholding the evidence _ and in contempt of court. The judge is holding hearings now in federal court in Columbus, Ga.
DuPont's zeal in defending Benlate claims has led to other allegations of misconduct.
In Florida, plaintiffs are alleging in state court that the company and its attorneys withheld test results in one trial two years ago and misrepresented the results of other tests in another case last year.
And in Hawaii, where DuPont fought unsuccessfully to the state's highest court to keep from turning over the initial Bush Ranch test results, a state court judge fined the company $1.5-million for intentionally withholding evidence. DuPont is appealing.
"The issues are complex enough without misconduct in discovery by counsel," Judge Ronald Ibarra said in his ruling in Kona, Hawaii.
DuPont says it has done nothing wrong. It contends that the final test results it presented in the Bush Ranch case had disproved the earlier findings and that its own test of Benlate had showed the fungicide was untainted.
DuPont's counsel, the prominent Atlanta firm of Alston & Bird, says in court pleadings that the tests presented were well founded and that its own conduct was "consistent with any relevant standard of professional responsibility."
The firm has also launched an aggressive counterattack, leveling charges of bias against the Bush Ranch judge and of misconduct against the plaintiffs' attorneys. DuPont says the plaintiffs settled because they had inside information that the jury was about to come back with a verdict for DuPont.
Rarely do aggressive defenses trigger the kind of charge DuPont faces _ of civil fraud. Indeed, in bringing such a complaint, the nursery plaintiffs are engaging in unusually aggressive tactics.
But the complaint poses a risk for DuPont as well as for Alston & Bird. A fraud finding could be used against the company in the 230 Benlate cases still pending; for the lawyers, it could result in disciplinary proceedings.
"The most powerful weapon that plaintiffs could come out of this with is a judicial finding of conscious wrongdoing by the defendant and counsel," says Stephen Gillers, a New York University law professor. "They could just take that into a courtroom and with very little more turn that into gold."
The Benlate nightmare began four years ago when a few nurseries that used it to keep plants fungus-free began complaining of stunted growth. DuPont pulled Benlate off the market amid concern that some might have gotten tainted with a common herbicide called Atrazine.
Soon, the complaints got worse. Big nurseries in Florida and Hawaii reported plants with a sickly yellowish cast, rotting leaves and dead roots. Fruit and vegetable growers told of acres of dead or deformed plants.
DuPont executives went to check. What they found, one of them jotted down on the flight back from Florida, was more serious than realized back at headquarters in Wilmington, Del.
"Many ornamental growers may be put out of business. . ." wrote Kent Reasons, director of development for agricultural products, in a memo that later became evidence in the Bush Ranch trial. "Lives are being shattered. . . . (We) may have only touched the tip of the problem."
The "greatest fear," Reasons noted, was that Benlate had been contaminated not by the common herbicide but by DuPont's new line of ultrapotent weedkillers, sulfonylureas, or SUs.
These were among the company's most promising new products. A high-profile judgment blaming them for crop losses would be a setback.
Circumstantial evidence pointed to the SUs. They and the active ingredient in Benlate were made at the same plant in Belle, W.
Va. And DuPont's quality control there and at other agricultural-chemical plants had been less than ideal, according to company records.
DuPont decided to start compensating growers. The reason: "overwhelming circumstantial evidence that Benlate is likely responsible for the damage in some way," according to a July 1991 internal memo now in court records.
But within months, the company and its attorneys started laying the groundwork for what would become its litigation strategy.
Its top outside lawyer for the Benlate problem was bracing for a wave of lawsuits. In a September 1991 letter to DuPont's legal department, Thomas M. Burke of the Orlando firm Cabaniss & Burke warned that finding the root cause could backfire in court.
Continuing research "at some level" made good "litigation sense," he advised. It would allow DuPont to argue that "it continues to search for a cause."
Burke, no longer DuPont's national coordinating counsel for Benlate, declines to comment. A DuPont spokesman says his advice was "speculative" and the company didn't follow it.
"We paid money when we thought Benlate was a problem," says the spokesman, Stacey J. Mobley. Payments stopped, he adds, after field testing showed they were no longer warranted. (DuPont announced in November 1992 that eight months of field testing had proved Benlate blameless and said it would no longer compensate growers.)
The huge liability potential was evident early on. New damage claims alleged yearly declines in nursery production, not just a one-year loss. And the claims were likely to exceed insurance coverage.
By the end of 1991, DuPont would take a $216-million charge for recalling Benlate and for early claims. It would also accrue another $343-million in Benlate costs in excess of its insurance coverage.
DuPont's Mobley says the growing liability didn't affect the decision to prepare for litigation. But those preparations nonetheless began almost immediately.
The Bush Ranch suit was filed in March 1992 by four nursery owners in Georgia, Alabama, Michigan and Hawaii.
It was assigned to Judge J. Robert Elliott, now 85.
Representing the growers was C. Neal Pope of Columbus, a veteran products-liability plaintiffs' lawyer and former Marine lieutenant.
Pope homed in on the potent SU herbicides as the most likely Benlate contaminant.
However, as Pope sought access to tests and other records, he ran into a wall.
The judge accused DuPont of stonewalling "under the guise of attorney-client privilege and work-product protection." DuPont's attorneys had invoked the privilege for more than 1-million documents without inspecting them to make sure it applied, the judge found.
In December 1992, an irritated Judge Elliott threatened to hold DuPont in contempt of court unless its chairman, Edgar S. Woolard, certified that the test results and claims reports had been turned over. They ultimately were.
But by then DuPont was committed to an all-out legal defense of Benlate. In announcing that it would stop compensating growers _ after having paid out $496-million in claims _ DuPont vice president William F. Kirk said, "We still do not know what caused the various crop losses, but we are now convinced it was not Benlate."
By the end of 1992, DuPont was projecting the cost of its litigation effort at $170-million, including $30-million just for legal fees, its records show. Bush Ranch was scheduled for trial in the summer of 1993.
As the trial date approached, DuPont continued to delay. Judge Elliott assessed a $500,000 fine _ to be canceled if it met his orders _ after finding that DuPont had "engaged in a continuous scheme and pattern of bad faith and discovery abuse." Two months later he raised the fine to $1-million.
Pope believed he had a strong circumstantial case. The only missing link, in his view, was any evidence of SUs in the soil.
For tests that the two sides agreed should be done, DuPont retained Alta Analytical Laboratories Inc. of El Dorado Hills, Calif. Three weeks before trial, it rushed 116 samples taken from the plaintiffs' properties. The results were to be shared with the plaintiffs.
Alta phoned the early results to Alston & Bird lawyers after four days of testing. The first two runs of 12 samples had picked up suspected traces of SUs and one positive hit above the minimum threshold in place for reliable results, according to Alta's records.
Three Alston & Bird lawyers listened as the numbers were read over the phone by Alta's principal scientists, Robert Bethem and Robert Peterson.
Peterson jotted down what the lawyers instructed next: "They want us to go back (after most of the work is done) and try to confirm (defirm) suspected positives."
Peterson has since said that the word "defirm" was his own, and both researchers maintain that the lawyers never pressured them to tailor the results.
Testing of all 116 samples continued for the next three days. Samples that yielded traces of SUs below the threshold were thrown out without further testing. The tests turned up 16 new hits, involving all four plaintiffs' properties, according to court records.
It was now two weeks before trial and only four days before Pope was scheduled to depose DuPont's expert witness on the cost of cleaning up contaminated soil.
DuPont was contending that because there was no contamination, there was nothing to clean up. But now, its own test results were possible evidence to the contrary.
What happened next isn't clear from court records, but as Bethem was reviewing the numbers by phone with Alston & Bird lawyers, he doubled the threshold for a positive finding.
He says that the decision was his own and that he had planned all along to use the higher threshold. The earlier suspected positives, he says, were too inconsistent to be reliable.
The higher threshold still didn't eliminate all positive findings. Four remained. One soil sample had tested positive three times. All four were now retested three more times, Bethem says, and came up clean each time.
But only the last set, showing "no detects" for SUs and faxed to the law firm at 1:20 a.m., was turned over to Pope. He wasn't given the raw data showing suspected positive findings, nor was DuPont's own expert witness who would interpret the summaries at the trial.
Alta's Bethem says that he stands by the final results showing no SUs and that he had no control over how Alston & Bird made use of the data.
DuPont made much of the results. "I am absolutely confident there are no sulfonylureas on these properties," its expert on contamination assessment, Nicholas Albergo of Tampa, testified.
In his closing argument to the jury, DuPont's lead attorney, Dow N. Kirkpatrick II, came back to the test results time and again.
For Pope, DuPont's summary of the results was a severe blow.
Pope had offered to settle before the trial and been rebuffed. Now, he says, he tried again, seeking out DuPont's general counsel, John M. Schmutz. This time, DuPont was interested.
DuPont offered $4.25-million to the four Bush Ranch plaintiffs, who had sought $430-million.
Pope also agreed to settle nine of his 10 other cases, for an undisclosed sum. The 10th is pending. The fines against DuPont were dropped.
At a news briefing, DuPont's chairman hailed the settlement as a "victory for DuPont, our employees and our science."
Of the dozen or so cases that have gone to trial since Bush Ranch, DuPont has, until recently, won outright or been hit with relatively modest damages. It has either settled these cases after trial or appealed.
That, however, is beginning to change. Earlier this year, jurors in Kona, Hawaii, awarded two growers $23-million in actual and punitive damages after seeing the complete Alta test results and learning that DuPont had withheld them.
In March, a state court judge in Honolulu awarded $8.5-million to nine growers and ruled that Benlate was a defective product. In April, he tacked on $3.2-million in interest because of delays due to DuPont's tactics.
Plaintiffs who lost a Benlate case in Polk County last year have asked the judge to set aside the verdict, order a new trial and sanction DuPont. The grounds: withholding evidence.
Last week in Miami, an orchid grower who said Benlate killed her 21,000 plants won a $2.5-million verdict against DuPont.