The digital economy created a measurement problem that economists struggled to solve for years. When Facebook launched in early 2004, for example, millions of people began using the platform daily, yet traditional economic statistics recorded its contribution to consumer welfare as zero. Similarly, when apps like WhatsApp largely replaced conventional SMS texting, GDP fell because texting revenue disappeared, even though consumer welfare increased substantially through the free messaging service.

The same applied to Wikipedia, Google Maps, and countless other free digital services that transformed how people work, communicate, and access information. This measurement gap means policymakers and business leaders operate with incomplete data on economic performance and consumer welfare, potentially leading them to misunderstand the true state of economic progress.

The challenge of invisible economic value

Gross Domestic Product has served as the primary metric for economic performance since the 1930s, but its creator, Simon Kuznets, warned in 1934 that "The welfare of a nation can scarcely be inferred from a measure of [GDP]." The rise of digital goods exposes fundamental limitations in how economies track value creation. Free goods, despite generating substantial consumer benefits, appear in national accounts with a measured price of zero and hence a measured value of zero. New goods face similar problems because no observed price existed before their introduction. As digital services proliferated, the gap between measured economic activity and actual welfare widened, creating what some called the modern productivity paradox.

Learn more: Driving competitive advantage through intangible investments

The implications extended beyond academic debates. Businesses investing in digital transformation need accurate data about returns, while policymakers require reliable metrics to assess economic health and make informed decisions. Traditional GDP could entirely miss benefits from free goods or even show the wrong direction when free goods replace paid services.

A framework for measuring digital benefits

As digital goods proliferate, there is an increased risk of misunderstanding the economy – unless conventional metrics of value are updated, according to researchers from Stanford University, Carnegie Mellon University, the University of British Columbia, Copenhagen Business School, and UNSW Sydney, who developed a new framework to address these measurement challenges. Published in the American Economic Journal: Macroeconomics, the study by Stanford University Professor Erik Brynjolfsson, Carnegie Mellon University Assistant Professor Avinash Collis, University of British Columbia Professor W. Erwin Diewert, Copenhagen Business School Professor Felix Eggers, and UNSW Business School Professor Kevin Fox introduced 'GDP-B', a metric that quantifies benefits from new and free goods.

The framework extended standard economic theory by deriving explicit terms for contributions of these goods to welfare change. Rather than requiring statisticians to recalculate GDP with estimated reservation prices, the approach provided adjustment terms that could supplement existing measurements. The researchers defined the reservation price as the price so high that consumers would choose not to buy the product.

“A key context of this research is that economies have changed structure quite dramatically since the development of GDP. There is now more knowledge production, more digital goods, and more things for free. This has led to calls to abandon the standard methods of economic accounting, to go ’beyond GDP',” said Prof. Fox.

“An alternative is to think of extending GDP, or ’GDP and beyond’. GDP-B provides an innovative approach that better measures economic activity and welfare in a modern economy and can be implemented by national statistical offices without jettisoning standard economic statistics. Rather, it extends them.”

Testing the framework with real data

The research team conducted incentive-compatible discrete choice experiments to measure consumer valuations. These experiments made participants' choices consequential, as prior research showed, producing externally valid results. For the Facebook study, researchers recruited 2885 participants through an online panel provider during 2016-2017, using quotas to match US census data for representativeness.

Participants chose between keeping Facebook access or giving it up for one month in exchange for payment. The study randomly assigned participants to different price points ranging from $1 to $1000. Researchers informed participants that one in 200 would be randomly selected to fulfil their choice, and that Facebook usage would be monitored remotely for 30 days to verify compliance. This design ensured truth-telling was the optimal strategy.

The median willingness-to-accept payment for foregoing Facebook for one month was $42.17, translating to approximately $506 per year. The research found that Facebook contributed between $231 billion and $1,013 billion to welfare in the United States over the period from 2003 to 2017, depending on the method used to estimate the reservation price. This translated into an average contribution of 0.05 to 0.11 percentage points per year to GDP-B growth, a substantial addition given that it represented just one digital service.

The authors noted that their estimates were conservative. As they explained, the valuation reflected "the price for giving up the 2017 version of Facebook, which includes all its attributes at the time, including the specific content and other participants of the social network and not simply the intrinsic features of the software."

Beyond social media: Smartphone quality changes

The framework also addressed quality improvements in existing products. Smartphones absorbed functionality from cameras, GPS devices, landline phones, gaming consoles, and calculators, yet traditional statistics struggled to capture this value. Laboratory experiments in the Netherlands measured consumer valuations for smartphone cameras using an incentive-compatible lottery method. Researchers found the median willingness-to-accept for forgoing smartphone cameras was €68.13 per month, translating to over €800 annually, while manufacturing costs ranged from $20 to $35. This gap demonstrated how quality improvements created substantial unmeasured welfare gains.

Subscribe to BusinessThink for the latest research, analysis and insights from UNSW Business School

“Besides providing the first empirical evidence on the value of smartphone quality changes to consumers, it also provides a framework that can be more generally applied to products that are similarly affected by rapid quality changes,” Prof. Fox explained.

Implications for business and policy leaders

The research provided several practical insights for business professionals and policymakers. First, traditional productivity and GDP measurements significantly understated economic progress and consumer welfare in the digital economy. Businesses investing in digital services created more value than conventional metrics suggested. Second, the experimental approach provided a practical means of measuring consumer valuations when hedonic techniques proved inadequate for rapidly changing products.

The authors emphasised that their framework could extend beyond digital goods to conventional products and services, providing "a more comprehensive and meaningful measure of welfare changes." As the researchers concluded, GDP-B and related metrics could "clarify which goods and innovations are actually contributing the most to economic growth and well-being as the economy evolves."

For organisations navigating digital transformation, the research validated investments in free or low-cost services that generated substantial consumer value. For policymakers, it highlighted the need for updated economic metrics that captured the full scope of welfare changes in modern economies. The measurement gap was real, quantifiable, and consequential for understanding economic performance.