Working Paper: NBER ID: w2993
Authors: Catherine J. Morrison
Abstract: Measures of productivity growth typically include in the Productivity "residual" the impacts of subequilibrium from fixity of factors, costs of adjustment, returns to scale and markups. This paper proposes a general two part framework for adjusting the residual measure to take these impacts into account. Errors computing the weights on output and quasi-fixed input growth in traditional measures are first corrected for both primal- and Cost-side measures. Then the deviation of revenues from costs is used to decompose the full primal measure to identify the differential influences of technical change, utilization fluctuations, scale economies and price margins. Use of the framework is illustrated empirically for the U.S.,, Japanese and Canadian manufacturing sectors, using an econometric model that allows explicit incorporation and measurement of these influences. The adjusted measures show that a significant amount of cyclical and secular change in measured productivity growth can be attributed to production characteristics other than technical change, particularly scale economies.
Keywords: Productivity Growth; Scale Economies; Markups
JEL Codes: O47; D24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| traditional measures of productivity growth fail to account for the effects of subequilibrium and scale economies (O49) | biases in the estimated productivity residual (C51) |
| adjusting for subequilibrium factors and scale economies (F12) | more accurate picture of productivity dynamics (O49) |
| adjustments made to productivity growth measures smooth cyclical and secular fluctuations in productivity (O49) | traditional measures increasingly underestimate productivity growth during periods of rising markups (O49) |
| adjustments reveal a more accurate picture of productivity dynamics (O49) | productivity growth attributed to technical change can instead be explained by production characteristics such as scale economies and markup behavior (O49) |