Working Paper: NBER ID: w21558
Authors: Sebnem Kalemli-Ozcan; Bent Sorensen; Carolina Villegas-Sanchez; Vadym Volosovych; Sevcan Yesiltas
Abstract: We construct nationally representative firm-level longitudinal data for European countries using financial statements from the Orbis database. We validate our data by comparing its coverage and firm size distribution to official statistics. We showcase two applications to show the importance of firm representativeness in understanding macroeconomic outcomes. First, we show that small-and-medium-sized firms (SMEs) account for a large share of aggregate economic activity. Second, we document that firm representativeness is important for calculating industry concentration trends over time as the share of economic activity accounted by top firms in an industry changes with the firm samples used.
Keywords: firm-level data; Orbis database; SMEs; industry concentration; macroeconomics
JEL Codes: E0; F0; O1
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| representativeness of firm data (L20) | economic contributions of SMEs (E69) |
| non-representative samples (C83) | misleading conclusions about increasing concentration trends (D30) |
| representative sample (C83) | declining trend in industry concentration (L49) |
| choice of accounting statements (M41) | concentration measures (D30) |
| representativeness of firms (L25) | reliable macroeconomic insights (E66) |