Working Paper: NBER ID: w30864
Authors: Christoph Boehm; Aaron B. Flaaen; Nitya Pandalainayar
Abstract: We measure export participation rates in the U.S. manufacturing sector using a new administrative dataset and compare them to participation rates constructed from the commonly used Census of Manufacturers (CM). Both at the establishment and firm level export participation rates are near 40 percent in the administrative data, almost twice as high as in the CM. The discrepancy appears to result predominantly from under-counting of small exporters in the CM. Our findings call for reconsidering the conventional wisdom that around 20 percent of manufacturing firms export.
Keywords: No keywords provided
JEL Codes: F0; F1; F14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| LFTTD dataset (Y10) | more accurate estimate of exporting firms (F14) |
| undercounting small exporters in CM (F10) | systematic underreporting (P37) |
| misreporting by survey respondents (C83) | systematic underreporting (P37) |
| exporting activities via intermediaries (F10) | systematic underreporting (P37) |
| CM methodology (C68) | respondent bias (C83) |