Can we Limit Algorithmic Coordination?

By: Michal Gal (ProMarket/University of Haifa)

Over the past years, our understanding of algorithmic harms in the marketplace has made significant strides. Nowadays, there is widespread consensus that algorithms not only have the potential to aid firms in establishing cartels but also possess the ability to independently learn and coordinate prices at supra-competitive levels, a phenomenon known as algorithmic coordination. This realization is supported by theoretical, experimental, and empirical research.

For instance, a study focused on the German gasoline market revealed that the transition from manual to algorithmic pricing by two firms in a duopoly led to a substantial increase in prices (9-28%). These collective findings leave us with an undeniable and credible conclusion: in specific market conditions, pricing algorithms can attain coordination at supra-competitive prices without any human intervention or pre-existing agreements, driven by their pursuit of a particular goal, such as “maximizing the seller’s profits.”

As we move forward, the next challenge is to effectively address the negative welfare impacts of algorithmic coordination without unnecessarily restricting suppliers’ ability to harness the advantages offered by algorithms. This task becomes even more crucial due to the exponential growth in the utilization of pricing algorithms in various industries…

CONTINUE READING…