Working Paper: CEPR ID: DP18112
Authors: Christos Genakos; Andreas Lamprinidis; James Walker
Abstract: This paper proposes a new algorithm with which to identify the potential effect of mergers by comparing the outcomes of interest in areas of overlap for the merging parties vis-à-vis areas of no overlap within a difference-in-differences estimation framework. Utilizing our proposed algorithm enables researchers and policymakers to perform retrospective merger evaluation studies that look at the effects of mergers on both price and non-price aspects. We demonstrate the applicability and value of our proposed methodology by examining the effects on price and product variety of four mergers of the late 1980s and the 1990s on the U.K. car market.
Keywords: automobile industry; merger analysis; policy evaluation
JEL Codes: L0; L1; L4; L5
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| mergers (G34) | market prices (P22) |
| Ford-Jaguar merger (F23) | market prices (P22) |
| BMW-Rover merger (G34) | Rover's prices (R48) |
| BMW-Rover merger (G34) | competitors' prices (D41) |
| Ford-Mazda merger (L23) | market prices (P22) |
| mergers (G34) | product variety (L15) |
| Ford-Mazda merger (L23) | product variety (L15) |