Analysis: Why is there a risk when algorithms set the prices you see

Discouraged by the £260 price tag for a train ticket to cover the 400km from Sunderland to London, an English fan found a cheaper way to get to Wembley Stadium to cheer for his team: traveling via Spain.

A Ryanair flight to the Spanish island of Menorca, an overnight stay in an airport hotel, a return flight to London and a car ride home all cost less than the extortionate price of a train.

The experience of this Black Cats fan [apelido do time de futebol Sunderland] it exemplarily reflects the costs, benefits, and occasional absurdities of algorithmic pricing, now employed by most rail and aviation companies to maximize their revenue.

These programs allow companies to raise prices quickly to exploit spikes in demand, for example when many football fans want to reach a stadium in another city.

But they can also automatically lower prices to attract customers when market conditions are weak, as with budget airlines.

At a time when consumers are concerned about rapidly rising inflation, pricing mechanisms need to come under stronger scrutiny. But digital markets present a special challenge, because many of the prices are determined by algorithms rather than by slow human beings.

In addition, service speed and flexibility can often be more decisive when shopping online than prices.

Why dynamic pricing can lead to more inflation

The promise of dynamic prices is that they match supply and demand better, generating greater economic efficiency. But they should certainly be attracting more attention from regulatory authorities than is currently the case. There are growing indications that algorithmic pricing can lead to unfair discrimination and encourage implicit collusion between companies, which would cause prices to rise overall.

The positive argument in defense of dynamic pricing is persuasively expounded by Yossi Cohen, who sells this type of service.

The Israeli founder of Quicklizard explains how his company combines historical data, product inventories and cost structures with external variables such as competitor prices, economic indicators and even the weather to automate pricing decisions. “We help companies find the sweet spot on the demand curve,” said Cohen.

Using dynamic pricing, developers say, can boost a retail group’s profits by between 7% and 10%, which can make a critical difference in an industry known for tiny margins.

But he argues that the ubiquity of price comparison services also increases the power of consumers to get the best bargains. “Online commerce has shifted the balance in favor of the consumer,” says Cohen.

Intuition might suggest that an online retailer’s ability to automatically track and respond to price cuts by rivals should intensify competition. But two studies, of which Alexander MacKay, an assistant professor at the Harvard School of Business Administration, is one of the authors, conclude the opposite.

How to regulate algorithms to avoid distortions

By studying how dynamic pricing works in the real world, MacKay found that when a company with a sophisticated pricing algorithm instantly tracks rivals’ price cuts, the price of products rises over time. A competitor with inferior technology would have no incentive to reduce prices. “This eliminates price competition from the game,” says MacKay.

Even if there is no explicit price collusion that attracts the attention of antitrust authorities, dynamic prices can weaken competition.

Smart and time-savvy shoppers may be able to do more research and take advantage of short-term deals and promotions, but other people won’t be able to do the same. “What’s so beautiful about competition is that when it’s strong, consumers don’t have to be especially sophisticated to benefit from it,” says MacKay.

In the past, regulatory authorities have intervened to correct abuses when innovations have emerged in pricing and in terms of market power. So, for example, in many countries retail is prohibited from selling products below cost in order to drive competitors out of business.

But the challenge of regulating online marketplaces is immensely complex given their speed and opacity. Businesses can also personalize offers, which means that two different consumers won’t necessarily see the same offer.

History indicates that general price controls are a bad idea. MacKay and his coauthors offer two more innovative solutions.

One would be to limit how often online companies can change their prices — e-commerce giant Amazon, for example, changes prices 2.5 million times a day across all of its product lines.

The second proposal would prohibit companies from including competitors’ prices in their logarithmic models. Both measures would encourage more blind competition and discourage implicit collusion.

In the absence of such regulatory intervention, consumers will have to rely on their smarts to navigate digital markets. The problem is, not all of us are as resourceful as that Sunderland fan.

Paulo Migliacci Translation