This paper, while simple, has some significant (and charged) conclusions. They purport to demonstrate that: (i) merging parties always decrease their innovation efforts post-merger while outsiders to the merger respond by increasing their effort; (ii) a merger tends to reduce overall innovation; (iii) consumers are always worse off after a merger; (iv) the model calls into question the applicability of the ‘‘inverted-U’’ relationship between innovation and competition to a merger setting.
The argument goes as follows:
- A merger between competitors affects the incentives to innovate through two channels: (i) the first channel relates to the (negative) externality that innovation by one firm has on its rival firms. A merger allows the merging parties to partially internalize this innovation externality and thus it lowers the incentives to innovate for the merged firm; (ii) the second channel relates to product market competition. This is relaxed after the merger so that profits increase both when firms do and do not innovate.
- A highly stylized model of a merger in an industry where innovation plays a key role is developed, that incorporates these two channels. The model considers the case of stochastic product innovation (as opposed to process innovation). In the model, absent innovation firms do not make any profits. This implies that innovation does not cannibalize any pre-existing profits and that the second channel can only act to promote innovation. Furthermore, there are no merger-induced efficiencies in the model, so that the effects on innovation incentives are based entirely on changes in competition between the merging parties.
- Many equations are then deployed. Instead of describing them (as if I could…), it is more important to enumerate the assumptions: (i) innovation is for the same homogenous product; (ii) likelihood of success is proportional to effort/investment put into innovation; (iii) pay-offs – if a firm did not discover the product, it gets a zero payoff. If only one firm was successful, it gets a prize normalized to 1. If two competing firms successfully discovered the product, each gets a prize worth δ. If three or more competing firms successfully introduce the product, it is assumed that competition upon the commercialization of the discovery is so strong that all firms get a zero payoff (i.e. we are looking at Bertrand competition here); and (iv) whenever both merging companies are successful in discovery, the merged company perfectly coordinates its price decisions between its two (identical) products.
- The conclusions are that: (i) since efforts to innovate are strategic substitutes, the merger will decrease efforts of the merging parties and increase the efforts of third parties; (ii) overall efforts to innovate will decrease following a merger in a concentrated industry, while the opposite holds true in a fragmented industry; (iii) the merger will be profitable given the possibility of the merging parties to coordinate their prices following the merger; (iv) given this price coordination, the classic price coordination effect of the merging parties is to reduce consumer surplus after the merger.
I understand equations are useful in formalising results and ensuring logical coherence. But I can also see that, given the assumptions deployed here, any further market concentration can only be bad for consumers.