Data is now part of the deal in every online transaction. Here's why we need to unbundle it

Data is often described as “the new oil,” the critical fuel for artificial intelligence and productivity growth. Governments debate how to regulate it. Firms compete to collect it. Investors treat companies as if their data were gold.

Yet for all this attention, one feature of data remains largely absent from policy discussions: its price. Unlike labour or capital, data rarely appears in national accounts or on balance sheets. For most citizens, it does not appear to have a price at all.

This invisibility matters. It reflects a structural feature of the digital economy that quietly shapes market power, consumer welfare and the distribution of economic gains. This invisible exchange is the key to unlocking consumer gains in the new data economy.

Unlike natural resources, data is rarely produced intentionally. It emerges as a by-product of economic activity. Every digital payment, blog post, online search or GPS-enabled commute generates information that firms capture.

From a policy perspective, this unusual production process for data creates a valuation problem. Traditional economic measurement relies on either observable market prices or production costs. Data has neither. Consumers are not explicitly paid when they generate it, and firms do not record a purchase when they acquire it. The data appears to be free.

Yet, firms invest heavily in extracting value from data. They build data centres, employ skilled labour and develop sophisticated algorithms. These investments signal that data is productive and scarce. Still, the underlying asset remains off the books.

This disconnect complicates policymaking. It obscures the true sources of value creation and makes it harder to assess market concentration, design consistent taxes and measure consumer surplus in data-intensive sectors.

Data is not actually free. Consumers are compensated, implicitly and invisibly. When firms derive value from transaction data, they have a clear incentive to stimulate more transactions. The way to stimulate more transactions is to lower prices. The result is a pervasive but largely unseen bargain. Consumers receive lower prices or free digital services. Firms receive both revenue and data.

Every digital interaction is therefore a bundled transaction: a purchase of goods or services, combined with a sale of data rights. The observed price paid is the net outcome of these two exchanges. The problem is that the two sides of the exchange are not separately observable: We do not know what price firms would have charged, if they could not harvest the data, and we do not know the difference between the price with data and the price without. This difference in prices is the unobserved price of data.

The bargain is usually good for consumers, but it could be much better. Consumers rarely observe the price they get for their data. They cannot choose to buy a product without selling information about themselves. And they cannot compare offers based on how much data is being extracted.

Economists refer to this as bundling: forcing two transactions to occur together. Bundling is not always harmful, but when one component has no visible price, it becomes impossible for buyers to assess whether they are getting a fair deal.

The result is a market in which consumers are perpetually naive sellers. Imagine arriving in a foreign country with an unfamiliar currency. On your first day, you overpay for lunch. But over time, you learn when to trade your currency for goods and when to search on. In digital markets, that learning never happens. Because data prices are never observed, every transaction feels like paying with the unknown currency on the first day.

This matters not only for individual welfare, but also for how the gains from data-driven growth are distributed. When the price of data is hidden, firms retain the upper hand. Consumers cannot demand better compensation for the information they provide, even as that information becomes more valuable.

There is a policy lever that could make data markets more transparent: unbundling. Requiring firms to offer a clear choice between transactions that allow data use and those that do not, at different prices if they so choose, would reveal the implicit value of data. The price difference would make the “data discount” observable.

Such transparency would empower consumers without mandating behavior. Some customers would choose lower prices, effectively selling their data. Others would pay more for privacy. For firms that find data privacy costly, they could change as high a price for such transactions as profitability requires. Over time, competition would determine market prices for data.

For regulators, unbundling would generate information. It would illuminate the value of consumer data, how that value differs across sectors and how it evolves over time. It would also enable more accurate accounting, competition analysis and policy design.

As leaders gather to discuss artificial intelligence, digital trade and inclusive growth, data valuation deserves a place on the agenda. The question is not whether data should change hands; it already does. The question is whether the rules governing its exchange are transparent and allow consumers to gain.

Making data measurable and its trade transparent would not resolve every issue of privacy, spillovers or market power. But when prices convey information, choices reveal preferences and policy can shape markets deliberately, this is a far better setting for consumers to thrive. Data and the artificial intelligence it enables will reshape the global economy. It may benefit consumers, as well as data-rich firms, but only if policymakers unveil the invisible deal at its core.