Data as Capital: Rethinking the Factors of Production

In previous generations, the economic factors of production consisted of land, labor, capital, and entrepreneurial ability.  Today, however, many businesses are conducted almost entirely online.  While they still need some amount of land, labor, capital, and entrepreneurial ability, they also need something else: data.  Few businesses can flourish in our digitized economy without key pieces of information to allow them to schedule, advertise, communicate, and collect payments.

Data as the New Capital

Traditionally, human laborers used pieces of physical capital goods (equipment) to do work and earn revenue.  Now, they need access to data to use digital tools to do work and earn revenue.  The more data, the more effective the worker, similar to how workers with more or better tools were more productive thirty years ago.  More data lets firms plan more, provide better customer service, and make more optimized spending and production choices.  

Essentially, data is a fifth factor of production that supplements modern, Internet-connected capital goods like computers.  Whereas entrepreneurial ability was once the “supplemental” factor of production that “supercharged” labor, data is the new “supplemental” factor of production that “supercharges” capital.  A firm may own dozens of modern computers and pay for high-speed Internet, but these tools are worth little without data sets that provide key information.

Like physical assets, such as capital goods, data can also be bought and sold in factor markets for resources.  Start-ups and established firms can purchase survey data and other market research to get insights into prospective customers.  They can also purchase data about rival firms, market trends, and contact information for prospective suppliers and regulators.  In a world where connections matter, data about these connections can be purchased.  If companies don’t spend on data and/or data analysis, they are largely operating blind.

Data and Firm Valuation

What is the value of data?  Some would consider data to be an intangible asset similar to a firm’s owned copyrights and other intellectual property.  These assets are typically harder to assign objective value than capital goods.  Some data, on its own, may be considered of little value.  Combined with other pieces of data, however, a data set may go from irrelevant to revenue-generating.  For example, data on total customer debt levels may not be highly relevant to a firm, but becomes far more valuable when combined with data about what type of debts each customer owes.  This would reveal that some customers are in better financial shape than initially appears, and thus can be marketed to.  

Some firms may collect and store data that is not particularly valuable to them, but would be highly valuable to other firms.  Additionally, some data may not be especially valuable in the short term, but becomes revenue-affecting in the long run when averaged with other data points.  As a result, firms have an incentive to collect and store as much data as possible, with its benefit potentially coming years later.  They may be able to market their vast data reserves as beneficial to potential buyers or mergers, arguing that future business leaders and managers could make use of such data to gain insights and boost productivity and revenue.  

This potential productivity benefit from strategic reserves of data is known as increasing returns to scale.  The more data you have, the more combinatory benefits you can enjoy, perhaps even unexpectedly.  Combinatory theory is similar to the concept of hybrid vigor, which argues that inputs from different sources often combine in ways that are stronger than when inputs come from the same source.  Hybrid sources of data may yield unexpected discoveries with valuable correlations, such as changes in consumer demand based on weather patterns or trending news stories. 

AI Supercharges the Value of Data

Until recently, huge data sets were likely undervalued due to lack of ability to analyze them effectively.  Today, the AI revolution is poised to make revenue-generating connections from mass data sets in ways never before anticipated.  Two variables that may have previously seemed unconnected can be linked by the supercomputing power of AI to reveal some revenue-bearing insight for a firm.  

Firms with massive data sets will have an incentive to purchase commercial AI software services to analyze them, creating new markets for commerce-related data analysis.  As these services prove successful at helping firms increase their revenue and profit through data insights, more firms will follow suit.  On a macro level, the use of AI to analyze years of data sets to find revenue-generating insights will likely lead to greater productivity and increased output.  Ultimately, faster economic growth will occur, fueled by ample data that can be analyzed and used for optimum productivity and sale-generating.

Ethics of Mass Data Analysis by AI

The petabytes of data waiting to be analyzed by commerce-focused AI likely contains plenty of private information.  While using data analysis to optimize production is unlikely to be controversial, using data analysis to optimize sales through customer analysis could violate some ethical norms.  Do customers want to be profiled by AI using years of their data, with such analysis subjecting them to targeted advertising?  Critics of AI might argue that such sophisticated analysis and targeted marketing is akin to addictive chemicals in products - too strong for the average consumer to resist.

Some such analysis and targeted marketing already exists, with online retailers like Target and Walmart using customer purchase histories and browsing data to target them with ads and suggested purchases.  Critics posit that this is fueling shopping addiction, with online customers captivated by continuous suggestions of products that they will almost certainly enjoy.  Policymakers are taking note of the potential harms of AI being unleashed on customers and targeting them for maximum revenue. 

Already, some governments are passing laws ensuring that businesses cannot use AI unethically without being held civilly liable.  These laws were passed in response to businesses claiming that AI was acting autonomously, trying to push responsibility onto the external software.  Basically, if your firm uses AI, your firm will be legally responsible for the content, such as targeted advertisements, that the AI creates on your behalf.

In regard to being a factor of production, therefore, governments are wanting firms to take responsibility for their data and AI usage as true employees and not as “independent contractors.”