OxyContin first sank its teeth into the poorer communities. “From a sales perspective, OxyContin had its greatest early success in rural, small town America—already full of shuttered factories and Dollar General stores,” Beth Macy wrote in Dopesick, her powerful account of opioid addiction and abuse. Within the first two years of the drug’s release, 24 percent of high school juniors in a small western Virginia town said they had tried OxyContin along with 9 percent of seventh graders.

In one slide, McKinsey told Allstate to try to settle 90 percent of its claims as quickly and as cheaply as possible. For the other 10 percent, policyholders or third-party claimants who didn’t take the Allstate offer or, even worse, hired a lawyer, the “boxing gloves” treatment was in order. They would fight in courts, for years if necessary, wearing down anyone who dared to sue.

McKinsey designed a system—the Claims Core Process Redesign—that pushed adjusters to make quick, lowball offers rather than allow them to come up with settlements that they considered fair. Adjusters, now tethered to a computerized claims system called Colossus, were reduced to little more than call-center workers reading prepared scripts. Pop-outs became rare. For homeowners’ claims, it was another computer program—Xactimate. But the idea was the same. Push claimants to accept less than the covered amount. Allstate says this characterization is “false and misleading” and that it overhauled its claims system in the 1990s to “pay claims more promptly and accurately.” McKinsey declined to comment.

By the 1990s, with McKinsey-led financialization sweeping the economy and ever-increasing pressure from activist shareholders for companies to boost profits, the firm pushed a big new idea to its clients: reducing the amount paid out in claims. In McKinsey-speak: “After years of squeezing the cost side, management recognized huge opportunities to rebalance and invested cautiously in LAE to capture indemnity savings.” The new approach to boosting profit was to curtail what insurance companies saw as unjustifiably high amounts paid out to some claimants. To control what it called “leakage.”

The surge in Allstate’s share price was accompanied by a dramatic fall in the “pure loss ratio,” the measure of claims payouts divided by premium income, before factoring in operating costs. In 1987, Allstate paid out 70.9 cents in claims for every dollar it took in. By 1997, two full years into the McKinsey makeover, the ratio had fallen to 58.2. By 2006, after spiking a year earlier amid huge claims resulting from Hurricane Katrina, it was 47.6.

The insurance adjuster’s role had been greatly diminished. During her father’s time, setting a value for a claim had been a skill. Under Allstate’s new system, the computer would spit out an estimate. This was Colossus, a program that analyzed hundreds of different injuries entered by adjusters like her. As Brady soon learned, Colossus had been tweaked to lowball claims amounts. It was then her job to persuade the policyholder to accept a claim even lower than the one disgorged by Colossus.

Mark Romano, one of the people responsible for tweaking Colossus so that it favored the insurer, worked out of Allstate’s headquarters in Northbrook, Illinois. “I could turn the knob, so to speak, and increase the values of cases, or I could turn the knob down and decrease the value,” he said.

When payout data indicated something was amiss, Romano took to the road to find out what was going on. He and his colleagues discovered that adjusters were entering an unusual amount of herniated disk injuries into Colossus, a far more valuable claim than a “soft tissue” claim like whiplash. Romano was told to “calibrate” the claims teams across the country, telling them to reduce the number of herniated disk claims “regardless if a neurologist or an orthopedic surgeon or a radiologist or whomever said it was a herniated disk.”

Skilling infused Enron with ideas he had learned at McKinsey, including the importance of periodically culling the herd. Either advance in the firm or leave. McKinsey called it “up or out.” At Enron it was “rank and yank.” McKinsey validated Enron’s strategy, including risk taking, securitizing loans to gas purchasers, and its “asset light” approach. As the McKinsey Quarterly explained, Enron became a world leader in private power generation “because it saw that profit did not depend on construction and operation skills, but on deal structuring and risk allocation.”

According to QuantumBlack’s home page, it helped a soccer team assess the “health of its players and identify the physical metrics that might signal impending injuries.” QuantumBlack’s evaluations were so detailed they included taking saliva samples. “Using objective medical markers and information from prior injuries, we identified the features that correlate to injury onset in the hamstring, upper leg, and lower leg,” the company said. As if to differentiate its work from the speculative chatter of sports radio and TV analysts, the company reported that its blind historical testing “correctly forecast 170 out of 184 non-impact muscle injuries.”

Some McKinsey consultants, especially younger ones, were bothered that the firm would so willingly help the House of Saud, a family from Riyadh whose patriarch conquered much of the Arabian Peninsula in the 1920s and whose sons have overseen a ruthless theocratic autocracy ever since. They argued that the firm should reduce, or possibly even end, its work there. Said one, “Saudi Arabia is a country that shouldn’t exist.”

The people who remained would have to work harder. McKinsey calculated that 1.7 percent of a doctor’s time was lost to tea breaks and said £400 million in savings could be realized if weak medical providers “achieve standard performance.”

But slashing jobs wasn’t enough. McKinsey also called for the end of “low value added healthcare interventions.” Translation: cutting back on what McKinsey deemed unnecessary medical procedures. For example, reducing certain hysterectomies by 70 percent could yield £80.6 million in savings; another £118 million could be saved by cutting knee joint surgeries by 30 percent.