One company choosing to use a quick and reactive pricing algorithm could lead its competitors to increase prices, leading to customers facing higher costs across the board, according to a recent study led by a University of Virginia economist.
Alexander MacKay, an associate professor of economics at the University of Virginia, said the research published by the National Bureau of Economic Research reviewed a concept called algorithmic coercion.
It’s what could happen when an algorithm leads an entire market to increase its prices for the same items, discouraging rival companies from trying to compete based on price and making mundane goods universally more expensive.
Airlines and hotels have used some variation of pricing algorithms for years, MacKay said, but “one of the big changes is monitoring rivals’ prices and reacting to that in real time.”
Online retailers are using software that monitors prices on competitors’ websites, and then they’ll change their prices in response, motivated to beat the prices they encounter.
“The role of economic theory and the research that we do is to look at, ‘well, what’s the implication of this for the consumer?'” MacKay said. “And what we’re pointing out is that this could actually lead to higher prices.”
Previously, traditional human pricing was used, allowing a person to set the price of an item. But pricing algorithms — formulas for setting prices based on inputted information — are becoming more common, MacKay said.
MacKay said if one company uses software that can collect a rival company’s prices quickly, and the technology can react in a fast way, “then that piece of software might be capable of disciplining any company that tries to lower their price.”
“And as a result, if the algorithm is powerful enough and the company is sort of large enough, it can really discourage any of its rivals from competing based on price,” MacKay said. “As a result, everyone’s going to set a much higher price.”
In some cases, MacKay said, the use of advanced pricing algorithms could result in prices that would be “higher than what you might get in a competitive market.”
“We also show in our paper that the prices could actually be so high that it would be worse for consumers than if the market participants got together and colluded on price,” he said. “So the potential of algorithmic coercion to raise prices is actually pretty substantial.”
MacKay said his research didn’t explore which companies are using the practice and what the impacts are, but it’s “quite possible that this is happening in a number of different industries.” He noted some online retailers and retail gasoline stations as specific examples.
