Working Paper: NBER ID: w30459
Authors: benjamin f jones; xiaojie liu
Abstract: Technological advance is often embodied in capital inputs. This paper develops a model where capital innovations occur on two margins: (1) vertically, where a capital input becomes more productive at a given task; and (2) horizontally, where a capital input replaces labor at a given task. These two forms of technological advance engage in a macroeconomic “tug of war” when capital and labor have less than a unitary elasticity of substitution, and the resulting framework can meet numerous macroeconomic regularities. First, it can produce a balanced growth path and satisfy the Uzawa Growth Theorem—even though all technological progress occurs in capital inputs. Second, it can produce intuitive macroeconomic dynamics, adding perspectives on the apparent productivity slowdown and declining labor share of income. Third, it can produce rich industry dynamics and inform structural change, including declining GDP shares of agriculture and manufacturing, sectoral bottlenecks, the role of general purpose technologies, and the limited macroeconomic impacts of computing. Overall, this tractable framework can help resolve puzzling tensions between micro-level observations of technological advance and macroeconomic features of economic growth.
Keywords: technological advance; capital-embodied technical change; economic growth; automation; productivity
JEL Codes: O10; O40
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| increases in automation (O31) | capital share in the economy (D33) |
| improvements in productivity of capital inputs (O49) | capital share in the economy (D33) |
| capital inputs (automation and productivity) (D24) | capital share (D33) |
| capital share in the economy (D33) | overall economic growth (O49) |
| automation of tasks and productivity improvements (O31) | balanced growth path (O40) |