This is another example of an early paper criticising the assumptions of those arguing for more stringent enforcement against mergers that may affect innovation – this time focusing on the potential for efficiencies brought about by such mergers.
While not new, the debate on the effect of mergers on innovation has been particularly lively in Europe since the European Commission’s broadened its innovation theory of harm, starting with Dow/DuPont. The reasons behind this debate lie in the opposite effects that mergers can have on firms’ incentives to invest in R&D. This paper, available here, argues that merger control should be a priori neutral as to the innovation effects of horizontal mergers, since the overall effect of a merger on innovation can be either positive or negative depending on the circumstances.
The paper further identifies a number of key factors which influence the impact of mergers on innovation. In particular, it suggests that a positive relationship between mergers and innovation is more likely for innovations that enhance demand (as opposed to innovations that increase price–cost margins) and increase the degree of horizontal differentiation. Positive effects also more likely if there are significant knowledge spillovers, or if merging firms can benefit from large complementarities by pooling R&D assets and coordinating research activities
Section II introduces the discussion on competition, mergers and innovation.
While some suggest that the analysis of unilateral innovation effects in merger cases bears resemblance with the analysis of unilateral price effects, the authors identify three fundamental differences between price and innovation effects. First, whereas the substitutability between competing products is sufficient to justify a presumption that the merged entity will raise prices absent efficiency gains, the existence of a potential negative innovation externality is not sufficient to make the claim that the merged entity will decrease its innovation in the absence of efficiency gains. The reason for this is that a merger may provide firms with higher incentives to innovate even if the merging firms’ investments in R&D exert negative effects on each other. For instance, if the merger leads to an increase in the merging firms’ price–cost margins, it may generate higher incentives to engage in demand-enhancing innovation that may outweigh the effect stemming from negative externalities between the merging firms (if such externalities exist). Second, R&D has a potential for spillovers which is absent in price effects. When spillovers exist, they may compensate partly, or even outweigh other negative externalities caused by the merger on incentives to innovate. Finally, whereas a static analysis of prices, such as the one performed in a unilateral price effects exercise, is meaningful, a static analysis of innovation is necessarily reductionist because of the fundamentally dynamic nature of the innovation process. In addition, R&D cannot be adjusted as quickly as prices, which might undermine merging firms’ ability to discontinue R&D (even when they have incentives to do so); further, standard unilateral effects analysis ignores the organization of R&D within the merged entity.
As a result, there is no consensus among economists about a presumed (negative or positive) sign of a horizontal merger’s impact on innovation. The literature shows that the impact of a merger is generally a combination of positive and negative effects. Leading economists do not agree about how to balance these effects in merger analysis. The literature, however, does provide guidance regarding the potential effects of a merger on innovation. This is discussed in the next section.
Section III examines the impact of a merger on firms’ incentives to innovate.
Mergers change firms’ incentives – e.g. merging parties have an incentive to cooperate, so the presumption regarding the unilateral effects of a merger on prices is that the changes in firms’ incentives will result in higher prices – but also their ability to produce. The new firm controlling all the merging firms’ assets can re-optimise the organisation of production. The reallocation of assets and/or reorganisation of production allows the merged entity to reduce marginal costs of production compared to the pre-merger situation. When innovation is factored in, the analysis becomes significantly more complex. Because innovation alters technologies, it is no longer possible to analyse the effect of new incentives resulting from the merger holding fixed production technologies and product lines. The analysis of incentives is complicated by the need to account for firms’ pricing behavior when analysing innovation effects. It is, therefore, necessary to understand how a merger affects incentives along two related dimensions—prices and investments in R&D.
If two companies investing in R&D merge, their inventions will compete with one another, thereby ‘diverting’ sales. Once the firms merge, they will have less incentive to invest and bring both innovations to the market (‘incentive diversion’). However, reduced rivalry at the research level (due to the removal of one competitor) can also induce firms to invest individually more than without the merger. First, the merged entity may also be tempted to reduce drastically its effort in one research path and increase substantially its effort in the other. Whether the merged entity will do so, or will instead reduce uniformly its efforts on both research paths, depends on the extent of decreasing returns in R&D. In any event, the resulting likelihood that an innovation will occur may be larger than in the no-merger scenario where both research paths are followed but with little effort (due to the risk of duplication). Second, there are several ways through which R&D allows a firm to make additional profits. To escape intense competition, a firm may invest in R&D to offer a product that is better than its rivals’, but it can also invest to propose a product that is different from its competitors’. Most rivalry in innovation combines both dimensions. A consequence is that R&D investment by one firm relaxes price competition and allows its competitor to sell more (potentially at higher prices). As a firm’s innovation raises rival’s demand, the innovation diversion effect is positive: it leads to an increase in the merging firms’ incentives to invest in R&D aimed at increasing horizontal differentiation. Third, when both merging companies innovate, the merged entity obtains two differentiated products. Since it can adjust prices to mitigate cannibalisation between them, the merged firm can obtain a higher profit with two innovations than with a single innovation.
Another important consideration is whether an innovation is demand-enhancement. Consider an innovator replacing an old product with a new one. The new product is of better quality and thus generates more sales. An innovator introducing a new product needs to decide how to balance its margin and sales so as to maximise profits. The key element in this trade-off is the own price-elasticity of demand: if the post-innovation demand faced by the innovator is significantly less elastic than pre-innovation demand, then the innovator will set the price of the new product above the price of the old product. However, if the post-innovation demand is as elastic as the pre-innovation demand, the innovator will prefer to maintain the price at its pre-innovation level, and sell higher quantities. It follows that both the demand expansion effect and the margin expansion effect – which has been the focus of papers finding a negative effect of mergers on competition – are essential effects that need to be considered, alongside the innovation diversion effect, when evaluating the likely impact of a merger on incentives to innovate. In general, the innovative firm will increase both margins and quantity so that both effects will coexist. These two effects may conflict, as the demand expansion effect is associated with a positive impact of a merger on innovation, whereas the margin expansion effect is associated with a negative impact on innovation. Which effect dominates depends on how innovation affects the level and elasticity of demand.
Yet another consideration when assessing the impact of a merger on innovation is technological spillovers, i.e. the “phenomenon that technological improvement by one company may help other companies improve their technology as well”. This includes imitation – which may be prevented by IP rights –, labour mobility across firms, scientific publications, etc. Such spillovers are positive innovation externalities and, therefore, their internalisation by the merged entity has a positive effect on its incentives to invest in R&D. Because spillovers affect firms’ incentives to innovate, they should be treated as part of the main competitive assessment conducted by competition authorities. The logic of the upward pricing pressure (UPP) methodology can be adapted to competition in innovation, and derives a simple formula for the “net innovation pressure” (NIP) that accounts for spillovers. An alternative is to replace the innovation diversion ratio with a “spillover-adjusted innovation diversion ratio” whose sign is given by the difference between the spillover ratio and the diversion ratio.
Section IV analyses the effect of complementarities between R&D assets and non-R&D-related cost reductions induced by a merger on innovation.
A merger fosters interactions between researchers with different experiences. By pooling talents from two research entities, it may induce creative emulation and foster new innovative ideas. Moreover, by reallocating talent towards the most promising research projects, a merged entity may raise its research productivity. Pooling of R&D assets de facto reduces production costs or raises the quality of the products offered to consumers. It occurs, for instance, when a process innovation by one merging partner reduces the production cost of the other partner. Moreover, pooling fosters the merged entity’s returns to R&D spending and, therefore, the merged firm’s investment in R&D. Efficiency gains at the production stage also matter for the impact of mergers on product innovation. The demand expansion effect relies on the marginal benefit from engaging in demand-enhancing innovation being greater the larger the firm’s price–cost margin. An increase in the margin can be generated by an increase of prices, but it can also result from a decrease in (marginal) production costs due to efficiency gains in production.
Section V concludes.
The academic literature on mergers and innovation does not support a presumption of negative impact of mergers on innovation. This conclusion follows from the existence of potential positive effects of mergers on innovation, even in the absence of spillover and complementarities. The authors discuss three main effects of mergers on the incentives to innovate: the innovation diversion effect, the demand expansion effect, and the margin expansion effect. Whereas the last one provides the merged entity with lower incentives to innovate, the second one provides it with higher incentives to innovate, and the sign of the first one depends on the nature of innovation.
Competition authorities should take a neutral perspective when assessing the impact of a merger on innovation, and should balance the various effects at work. All the effects of a merger on the incentives to innovate identified in this paper, including spillover effects, should be part of the main competitive assessment carried out by competition authorities.
While taking the form of a literature review, this paper is hard going – a consequence of its efforts to refute literature that argues that the predominant effect of horizontal mergers on innovation is negative. This is at the heart of a number of imbalances – such as a single section taking over two-thirds of the paper, and the very short thrift given to arguments that horizontal mergers may be detrimental to innovation.
If one can look past this, however, one gets a fairly comprehensive discussion of the literature on the impact of mergers on innovation, coupled with a boilerplate message: ‘let’s take each case on its own merits’. What is missing, for a paper with such clear addressees, is a discussion of how best to identify those mergers where the risks or benefits to competition prevail, and whether some rules of thumb (or presumptions) would be beneficial to merger analysis.