A Model of the Data Economy

The paper builds a tractable dynamic general-equilibrium model in which data is information that firms accumulate to forecast a moving target, is generated as a by-product of production, is tradeable and (partially) non-rival, and depreciates over time. Its primary stated contribution is the framework itself — a tool that maps to many observable macro and finance measures and can be “calibrated and estimated like its old-economy DSGE counterpart” — rather than any single prediction. The model implies that a data economy resembles a Solow capital-accumulation economy with diminishing returns in the long run, but exhibits new short-run behavior: increasing returns from a “data feedback loop,” S-shaped firm growth with possible growth traps, initial losses, and the barter of data for goods. Because data can be exchanged at a zero monetary price, the authors argue conventional GDP can understate economic activity, and the model delivers a data-specific depreciation formula and a recursive value function for valuing data. In the baseline economy equilibrium choices are socially efficient despite non-rivalry, increasing returns, and data being a by-product; but once data is also used for “business stealing,” the model generates over-investment in capital and excessive trade in data. These are properties of a stylized model under specific assumptions (competitive price-taking firms, normally distributed shocks, a quadratic quality loss, an aggregate state variable, and an exogenous rental rate of capital), not calibrated empirical estimates.

Laura Veldkamp | Macro Paper Warehouse

A Model of the Data Economy