Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/171071 
Authors: 
Year of Publication: 
2017
Series/Report no.: 
CESifo Working Paper No. 6607
Publisher: 
Center for Economic Studies and ifo Institute (CESifo), Munich
Abstract: 
Mergers realize heterogeneous competitive effects on profits, production, and prices. To date, it is unclear whether differential merger outcomes are caused mostly by firms’ technology or product market attributes. Furthermore, empirical merger studies conventionally assume that, conditional on regressors, the impact of mergers on outcomes is the same for every firm. We allow the merger responses to vary across firms, even after controlling for regressors, and apply a random-coefficient or heterogeneous treatment effect model (in the context of Angrist and Krueger (1999), Heckman, Urzua, and Vytlacil (2006), and Cerulli (2012)). Based on a comprehensive dataset on the static random access memory industry, we find that firms’ postmerger output further increases (and postmerger price further declines) if merging firms are more efficient, operate in more elastic product markets, are more innovative, and acquire knowledge in technological areas that are relatively unexplored to themselves. A further interesting insight is that product market characteristics cause stronger postmerger outcome heterogeneities than do technology market characteristics. We also find that the postmerger effects accounting for heterogeneities differ greatly from those that consider homogeneous postmerger outcome effects. Our estimation results provide evidence that ignoring heterogeneous outcome effects can result in heterogeneity bias, just as ignoring premerger heterogeneities can lead to selectivity bias.
Subjects: 
heterogeneous treatment effects
horizontal mergers
market power effects
merger evaluation
premerger heterogeneity
postmerger heterogeneity
JEL: 
L11
L13
L52
O31
O32
O38
Document Type: 
Working Paper
Appears in Collections:

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.