Toward an Understanding of Learning by Doing: Evidence from an Automobile Assembly Plant

Working Paper: NBER ID: w18017

Authors: Steven D. Levitt; John A. List; Chad Syverson

Abstract: Productivity improvements within establishments (e.g., factories, mines, or retail stores) are an important source of aggregate productivity growth. Past research has documented that learning by doing-productivity improvements that occur in concert with production increases-is one source of such improvements. Yet little is known about the specific mechanisms through which such learning occurs. We address this question using extremely detailed data from an assembly plant of a major auto producer. Beyond showing that there is rapid learning by doing at the plant, we are able to pinpoint the processes by which these improvements have occurred.

Keywords: Learning by doing; Productivity; Automobile assembly; Defect rates

JEL Codes: D2; L2; L6; O3


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
cumulative production experience (E23)defect rate reductions (L15)
second shift (J29)defect rates (L15)
absenteeism (J22)defect rates (L15)
defect rates (L15)warranty payouts (G35)

Back to index