April 20, 2026

Vipin Arora Official Portrait

A blog post from BEA Director Vipin Arora

I’ve struggled over the years with some basic grammar (and I know I’m not the only one). It’s hard for me to remember what an adverb is, or to pick out a simile versus a metaphor. You can add treating the word data as plural to that list. Saying “the data are” instead of the “the data is” doesn’t sound right to me, even though I know that the singular usage has arisen only in recent years. And just like the categorization of data as singular or plural has evolved, so has the way data are treated in economic accounting.

Traditionally, statisticians have treated spending on the production of data for use within an organization as an expense when compiling gross domestic product. The latest international guidelines, however, call for treating spending on data production as an investment. This means certain types of data will be classified as a produced asset—just like a tractor. For example, the money that manufacturers spend to collect and store sensor data on their equipment will now be considered an investment, not an expense. That might not be intuitive, yet it makes sense when you think about two common characteristics that some produced data share with more traditional capital, such as tractors. 

First, like traditional capital, some produced data are used repeatedly or continuously in production processes for more than one year. Just as we would expect a farmer to use a tractor for many years in harvesting crops, we would also expect a manufacturer to use sensor data on equipment over several years. Sensor information gives manufacturers real-time data from production lines, helping to reduce unexpected downtimes (by predicting equipment failures and scheduling proactive maintenance), enhancing operational efficiency, and monitoring quality control. 

Second, both traditional capital and produced data that are used for more than a year provide economic benefits to the owner. Going back to our examples, both the tractor and sensor data are directly used to generate output—in one case it may be a farm’s crops and in the other it might be a manufacturer’s cars.

“Capitalizing” data—an improvement we plan to introduce with the next comprehensive update of GDP and its major components in 2028—has been on our radar for some time. BEA has been studying, testing, and improving its methods for treating data as an asset in the economic accounts for many years. Its inclusion will result in more accurate and relevant statistics and allow us to better measure the effects of innovation and intangible assets on economic growth and productivity.