Economists challenge the “data equals power” narrative.

Data has become a critical input into digital markets, but its effects are neither uniform nor predictable. This forces regulators and companies to confront a more complex set of trade-offs than traditional competition frameworks had anticipated.

That was the summary of the International Competition Policy Foundation’s interview, which was conducted by the PYMNTS-owned publication Alexandre de CornierProfessor of Economics and Director of the Center for Competition Policy and Regulation at the University of Toulouse School of Economicsand Greg TaylorProfessor of Digital Markets and Competition at Oxford Internet Institute, Oxford University.

Their research is detailed in “Data-driven competition: Implications for enforcement and merger control“, rejects the hypothesis that data accumulation alone determines market power.

How data is used determines competitive results

Taylor drew a clear line between two distinct economic influences. “Data is not a monolithic thing, but it can be used in many different ways, and in different types of business models,” he said.

He continued: “This data will make companies more able to generate income from their customers, and this makes each customer more valuable. This will motivate companies to compete more intensely to attract these valuable customers.”

However, he accompanied this with a second observation: that the data could also be used to make companies more efficient at “extracting value from these customers, for example by enabling price discrimination.”

This tension is at the heart of data-driven competition. When data is used to improve services, it can raise quality and intensify competition. When it is used to improve targeting or pricing, it can increase revenue without improving results for users.

The dynamic nature of data

De Cornier stressed the need to evaluate data over time. “Data has a dynamic dimension,” he said, explaining that more users today generate more data tomorrow, which then shapes future competition.

The dynamic quality of information flow supports the idea of ​​feedback loops. Companies that attract users collect more data, which can be used to improve products and attract additional users.

However, this episode is not guaranteed to promote dominance. Taylor explained the condition that determines whether it will be applied or not. “We need to believe that having better data allows companies to produce better products, and that brings more users, and with them more data.”

If this condition is met, the course will strengthen the company’s position. If not, the loop can stop. When data is used primarily to extract value from users rather than improve the product, “that’s not going to be something that will attract a lot of people to your products, and that will then break the cycle,” Taylor noted.

Positive feedback versus consumer harm

The distinction between product improvement and value extraction explains why data can produce sharply different results across markets.

When companies use data to improve recommendations, search results, or logistics, the result is often greater engagement and stronger competition for users. When companies use data to increase ad intensity or personalized pricing, the result may be higher revenues with little improvement in user experience.

Taylor pointed to measurable indicators as a way to differentiate between these outcomes. “If we believe that data is mostly used to train better search algorithms, we should be able to use engagement metrics to see if that is reflected in the way consumers interact with those algorithms,” he told CPI. “If the data is used to target ads, that’s something we can measure as well.”

For industry leaders, this framework links data strategy directly to observable results. It also indicates where regulatory scrutiny is likely to focus.

Harm theory chapter

The interaction between these influences increases the complexity of the implementation process. De Cornier cautioned against combining incompatible theories when assessing competitive harm.

“For the sake of consistency of argument, agencies should perhaps avoid pursuing theories of harm that are both exclusionary and exploitative,” he said.

The reason is structural. A company that uses data to improve products and attract users may build a lasting competitive position. A company that uses data to extract value risks undermining its growth by making its offering less attractive.

Thus, regulatory bodies face a choice. They must determine which mechanism works in a particular market and build their case accordingly.

Data trading and merger structure

These differences become more apparent in the analysis of mergers. The research highlights the role of data commerce, particularly whether companies can share or sell data before merging it.

De Cornier described one of the key findings: “The merger is likely to benefit consumers where data trading is hindered or impossible.”

In such cases, companies may not be able to monetize data across markets prior to the merger. Combining these companies could create incentives to improve products in order to collect more data, which could lead to increased consumer surplus.

When data can be traded, the risks change. Companies may use mergers to limit access to data, reducing competition in adjacent markets.

This puts data trading at the center of merger review. Regulators should examine whether companies exchange data before a deal is struck, whether they plan to do so, and why such exchanges might be restricted.

Evaluate when to proceed with the deal

The framework also requires regulators to examine the source of those restrictions. If the data cannot be shared due to privacy rules, allowing companies to combine the data through a merger may conflict with those protections. If limitations arise from formatting issues or concerns about misuse, mergers may enable the data to be used more efficiently.

De Cornier framed the issue as one of incentives. When companies cannot trade data, they may not fully understand its value across markets. A merger could change this calculus by aligning incentives and encouraging companies to improve their offerings to attract users.

The task for regulators is to determine which effect will dominate in each case. For companies, the task is to understand how data strategy impacts growth and auditing.

As De Cornier said, the logic ultimately depends on whether data leads companies to deliver better outcomes: “It needs more data to allow the company to offer a better deal to consumers.”