Working Paper: CEPR ID: DP16027
Authors: Ulrich Doraszelski; Jordi Jaumandreu
Abstract: De Loecker & Warzynski (2012) obtain the markup from the firm’s cost minimization problem by substituting in estimates of the output elasticity of a variable input and the disturbance that separates actual from planned output. We show, however, that consistently estimating the output elasticity and the disturbance using the procedure developed in Olley & Pakes (1996) and Levinsohn & Petrin (2003) generally requires observing and controlling for the markup. We analyze the resulting biases and discuss alternative approaches to estimation.
Keywords: No keywords provided
JEL Codes: No JEL codes provided
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| unobserved demand heterogeneity (R22) | biased estimates of firm-level markups (L11) |
| DLW method without accounting for demand differences (C69) | incorrect estimated markups (D43) |
| incorrect estimated markups (D43) | conclusion that acquiring firms have decreased their markups post-acquisition (G34) |
| bias in estimated markups (L11) | correlation with firm's export status (F10) |
| bias in estimated markups (L11) | correlation with trade liberalization measures (F13) |
| unobserved demand heterogeneity (R22) | estimated markup levels (D43) |
| unobserved demand heterogeneity (R22) | relationship with market dynamism (L14) |