An Electrifying Puzzle

Economist Lester Lave ’60 shows how energy traders cracked the prisoner’s dilemma

It’s a gloomy November afternoon, the sky is a menacing gray, and a relentless Portland drizzle is turning the canyon into a giant bowl of mud. Clutching my soggy notebook, I am struggling to keep pace with Lester Lave ‘60—and not just because of his brisk stride. Over the course of an hour-long interview, we have discussed photovoltaic cells, wind power, artificial intelligence, carbon trading, Caesar, football, smog, and the dynamics of shame as an economic force. As we stroll past Chittick, his freshman dorm, I keep wishing he’ll pause for a reminiscent reverie and give me a chance to catch up with my notes. It doesn’t happen. Standing still is not part of Lave’s repertoire.

Lave returned to campus last fall to give a lecture on electricity restructuring, a subject on which he is considered one of the nation’s leading thinkers. But that’s hardly his only field of expertise. In addition to holding the Harry B. and James H. Higgins chair of economics at Carnegie Mellon University in Pittsburgh, he is also director of the Green Design Institute; co-director of the Electricity Industry Center; member of the Institute of Medicine of the National Academy of Sciences; and a founding member of Pittsburgh’s Group Against Smog and Pollution (GASP). He has written papers on subjects as diverse as air pollution, ethanol, dams, and raisins. A recent paper was titled “Planning for Natural Disasters in a Stochastic World.”

“Lester is the quintessential Reedie,” says economist Seth Blumsack ’98, an assistant professor in the department of energy and mineral engineering at The Pennsylvania State University, who studied under Lave at CMU. “He is a very intense guy, clearly brilliant, passionate, a really good thinker, intellectually agile. He has tackled all sorts of problems that on the surface had nothing in common, but is able to apply insights in one area to others.”

Lave grew up in Santa Monica in the 1940s and ’50s, then moved to Manhattan Beach, a place he recalls as “intellectually barren.” When he arrived at Reed, Lave remembers, “I thought I had gone to heaven.” Finally, he found himself among intellectual peers. (Coincidentally, his freshman roommate was Howard Wolpe ’60, who later served seven terms as congressman in Michigan. Lave’s older brother, Charles Lave ’60, also attended Reed, and died last year.) Lave came intending to study physics, but fell in love with the classics. He read Caesar’s Gallic Wars in Latin. “I couldn’t go to sleep,” he says. “I read it all night.”

Soon Lave fell under the spell of what Thomas Carlyle dubbed “the dismal science.” Reed’s economics department then included several legendary professors, among them Carl Stevens ‘42, Art Leigh, and George Hay. Lave wrote his thesis for Hay on the prisoner’s dilemma—an infamous intellectual puzzle that would later prove pivotal in unraveling the California energy-trading scandal.

The classic form of the prisoner’s dilemma runs like this: two men are arrested and accused of a dreadful crime. However, the evidence against the two is flimsy, so the authorities offer each prisoner the following deal: if you confess and inform on your partner, you will spend only five years in prison. However, if you plead innocent, and your partner informs on you, you will spend your entire life behind bars. The prisoners are held in separate cells, and cannot communicate. If they both plead innocent, they will walk away scot-free—but neither can trust the other not to defect. In most scenarios, rational actors choose the “safe” course of minimizing their maximum punishment—and betray their partners.

The prisoner’s dilemma has applications in dozens of fields from evolutionary biology to nuclear disarmament to competitive cycling. It also applies to the electricity auctions held by the California regional transmission organization (RTO) following the energy deregulation of the late ’90s. Starting in 1998, the RTO held hourly auctions inviting generators to make bids for delivering a certain number of megawatts to the grid for 60 minutes. In each auction, the RTO would select the lowest bidders until it had accumulated enough power to meet projected demand.

In theory, generators in this scenario should keep their bids low for fear of being undercut and losing out—the logical equivalent of playing safe in a multi-player version of the prisoner’s dilemma. Of course, if the generators got together, they could secretly agree to raise their bids, and reap massive profits. But such collusion is totally illegal, and would run afoul of the Sherman Antitrust Act.

But what if a generator one day puts in a bid that appears, on its face, so high as to be nonsensical? The other generators would probably raise their eyebrows, put in lower bids, and win the contract for that hour. But then, perhaps, they might pause, and reflect that if they all raised their bids, they would all make more money. And, perhaps, in the next auction, other generators might signal their understanding by putting in their own ridiculous bids—the trader’s equivalent of a wink at the bridge table.

This is more or less what happened in California. Starting in Spring 2000, generators submitted progressively higher bids, driving the wholesale price of a kilowatt-hour from 2¢ up to $1. “People figured out how to beat the game,” Lave says.

By cracking the prisoner’s dilemma, generators made billions of dollars in profits (and wrought havoc on California’s economy). Much of this scandal can be blamed on unscrupulous traders, but the botched restructuring of the California market is also at fault. “Setting up the system in this way, without a regulator, was equivalent to telling the energy traders, ‘Please don’t cheat,’” he says.

Lave argues that the structure of the RTO auctions contributed to the problem by giving traders an easy way to signal one another. If the auctions were held, for example, only once a year, the cost to a generator of losing a bid would be much higher—high enough, perhaps to discourage signalling.

Perhaps more important, however, is his conclusion that deregulation for the sake of deregulation has not produced lower prices. “Blind faith is unlikely to produce a free market that is competitive,” he wrote in a recent paper. “Substituting markets for traditional regulation is only one choice among many policy instruments to achieve a goal of lower prices; such substitution should not be a goal in itself.”

In addition to his work on deregulation, Lave is also known for an influential 1970 paper in Science that calculated the social and economic costs of air pollution (then viewed as primarily an aesthetic issue). He also earned a certain notoriety—or, as he puts it, “a vast amount of animosity,” for his conclusion that the EV1, the electric vehicle produced by General Motors in the mid ’90s, would ultimately emit more air pollution than a traditional gasoline car because of the enormous quantity of lead in its 1,100-pound batteries.

“What I got from Reed was a certain kind of curmudgeonliness,” Lave says. “A desire to get to the bottom of an issue. Let’s not just salute the flag. Let’s figure out what’s real and what’s hype. It doesn’t win you a lot of friends—but it does lead you to a lot of interesting conclusions.”

—Chris Lydgate