Data, Intangible Capital, and Productivity
This paper analyzes how the increased use of data in economies affects productivity. We introduce a framework for measuring data and find that data assets are conceptually encompassed in the Corrado, Hulten, and Sichel (2005, 2009) intangible capital framework. Data assets are not explicitly identified in that framework, however, and to remedy this, the paper develops measures of industry-level investments in data for nine European countries. Analysis of the new measures concludes that about 50 percent of intangible capital is, in effect, data capital.
Next, a simple model of an economy with data/intangible capital is used to assess the impact of the increased use of data, especially proprietary bigdata. We find that there are two primary macroeconomic impacts. First, the greater relative efficiency of data capital boosts its contribution to labor productivity. Second, the increased data intensity of intangibles weakens knowledge diffusion and diminishes TFP growth. To frame the net magnitude of these offsetting effects, the paper uses the recently developed EUKLEMS & INTANProd database (LLEE 2023). The boost to labor productivity stemming from the estimated relative efficiency of data capital is offset by the appropriability effect, which shaved .3 and .4 percentage points off 2010-2019 TFP growth in Europe and the United States, respectively.