Working Paper: CEPR ID: DP7773
Authors: Julian Di Giovanni; Andrei A. Levchenko; Romain Rancière
Abstract: Existing estimates of power laws in firm size typically ignore the impact of international trade. Using a simple theoretical framework, we show that international trade systematically affects the distribution of firm size: the power law exponent among exporting firms should be strictly lower in absolute value than the power law exponent among non-exporting firms. We use a dataset of French firms to demonstrate that this prediction is strongly supported by the data. While estimates of power law exponents have been used to pin down parameters in theoretical and quantitative models, our analysis implies that the existing estimates are systematically lower than the true values. We propose two simple ways of estimating power law parameters that take explicit account of exporting behavior.
Keywords: firm size distribution; international trade; power laws
JEL Codes: F12; F15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| International trade (F19) | Distribution of firm size (L25) |
| Exporting behavior (F10) | Power law exponent among exporting firms (D22) |
| Firm size (L25) | Exporting likelihood (C59) |
| Entry of firms into international markets (F23) | Distribution of firm sizes (L25) |
| Failure to account for exporting behavior (F10) | Power law estimates (C51) |